#!/usr/bin/env python import sys, random, os, math, copy import option_list as OL, afni_util as UTIL import lib_rand_timing as LRT import lib_afni1D as LAD g_help_string = """ =========================================================================== Create random stimulus timing files. The object is to create a set of random stimulus timing files, suitable for use in 3dDeconvolve. These times will not be TR-locked (unless the user requests it). Stimulus presentation times will never overlap, though their responses can. --------------------------------------------------------------------------- ** There is now basic (old) and advanced usage. Until I decide whether and how to merge the help, consider: make_random_timing.py -help_advanced Otherwise, what follows is the basic usage. Though all options are listed here. --------------------------------------------------------------------------- This can easily be used to generate many sets of random timing files to test via "3dDeconvolve -nodata", in order to determine good timing, akin to what is done in HowTo #3 using RSFgen. Note that the -save_3dd_cmd can be used to create a sample "3dDeconvolve -nodata" script. given: num_stim - number of stimulus classes num_runs - number of runs num_reps - number of repetitions for each class (same each run) stim_dur - length of time for each stimulus, in seconds run_time - total amount of time, per run pre_stim_rest - time before any first stimulus (same each run) post_stim_rest - time after last stimulus (same each run) This program will create one timing file per stimulus class, num_runs lines long, with num_stim stimulus times per line. Time for rest will be run_time minus all stimulus time, and can be broken into pre_stim_rest, post_stim_rest and randomly distributed rest. Consider the sum, assuming num_reps and stim_dur are constant (per run and stimulus class). num_stim * num_reps * stim_dur (total stimulus duration for one run) + randomly distributed rest (surrounding stimuli) + pre_stim_rest + post_stim_rest (note: account for response time) ----------- = run_time Other controlling inputs include: across_runs - distribute num_reps across all runs, not per run min_rest - time of rest to immediately follow each stimulus (this is internally added to stim_dur) seed - optional random number seed t_gran - granularity of time, in seconds (default 0.1 seconds) tr_locked - make all timing locked with the accompanying TR The internal method used is similar to that of RSFgen. For a given run, a list of num_reps stimulus intervals for each stimulus class is generated (each interval is stim_dur seconds). Appended to this is a list of rest intervals (each of length t_gran seconds). This accounts for all time except for pre_stim_rest and post_stim_rest. This list (of numbers 0..num_stim, where 0 means rest) is then randomized. Timing comes from the result. Reading the list (still for a single run), times are accumulated, starting with pre_stim_rest seconds. As the list is read, a 0 means add t_gran seconds to the current time. A non-zero value means the given stimulus type occurred, so the current time goes into that stimulus file and the time is incremented by stim_dur seconds. * Note that stimulus times will never overlap, though response times can. * The following options can be specified as one value or as a list: -run_time : time for each run, or a list of run times -stim_dur : duration of all stimuli, or a list of every duration -num_reps : nreps for all stimuli, or a list of nreps for each Note that varying these parameters can lead to unbalanced designs. Use the list forms with caution. Currently, -pre_stim_rest and -post_stim_rest cannot vary over runs. ---------------------------------------- getting TR-locked timing If TR-locked timing is desired, it can be enforced with the -tr_locked option, along with which the user must specify "-tr TR". The effect is to force stim_dur and t_gran to be equal to (or a multiple of) the TR. It is illegal to use both -tr_locked and -t_gran (since -tr is used to set t_gran). ---------------------------------------- distributing stimuli across all runs at once (via -across_runs) The main described use is where there is a fixed number of stimulus events in each run, and of each type. The -num_reps option specifies that number (or those numbers). For example, if -num_reps is 8 and -num_runs is 4, each stimulus class would have 8 repetitions in each of the 4 runs (for a total of 32 repetitions). That changes if -across_runs is applied. With the addition of the -across_runs option, the meaning of -num_reps changes to be the total number of repetitions for each class across all runs, and the randomization changes to occur across all runs. So in the above example, with -num_reps equal to 8, 8 stimuli (of each class) will be distributed across 4 runs. The average number of repetitions per run would be 2. In such a case, note that it would be possible for some runs not to have any stimuli of a certain type. ---------------------------------------------------------------------- examples: 1. Create a timing file for a single stimulus class for a single run. The run will be 100 seconds long, with (at least) 10 seconds before the first stimulus. The stimulus will occur 20 times, and each lasts 1.5 seconds. The output will be written to 'stimesA_01.1D'. make_random_timing.py -num_stim 1 -num_runs 1 -run_time 100 \\ -stim_dur 1.5 -num_reps 20 -pre_stim_rest 10 -prefix stimesA 2. A typical example. Make timing files for 3 stim classes over 4 runs of 200 seconds. Every stimulus class will have 8 events per run, each lasting 3.5 seconds. Require 20 seconds of rest before the first stimulus in each run, as well as after the last. Also, add labels for the 3 stimulus classes: houses, faces, donuts. They will be appended to the respective filenames. And finally, display timing statistics for the user. The output will be written to stimesB_01.houses.1D, etc. make_random_timing.py -num_stim 3 -num_runs 4 -run_time 200 \\ -stim_dur 3.5 -num_reps 8 -prefix stimesB \\ -pre_stim_rest 20 -post_stim_rest 20 \\ -stim_labels houses faces donuts \\ -show_timing_stats Consider adding the -save_3dd_cmd option. 3. Distribute stimuli over all runs at once. Similar to #2, but distribute the 8 events per class over all 4 runs. In #2, each stim class has 8 events per run (so 24 total events). Here each stim class has a total of 8 events. Just add -across_runs. make_random_timing.py -num_stim 3 -num_runs 4 -run_time 200 \\ -stim_dur 3.5 -num_reps 8 -prefix stimesC \\ -pre_stim_rest 20 -post_stim_rest 20 \\ -across_runs -stim_labels houses faces donuts 4. TR-locked example. Similar to #2, but make the stimuli TR-locked. Set the TR to 2.0 seconds, along with the length of each stimulus event. This adds options -tr_locked and -tr, and requires -stim_dur to be a multiple (or equal to) the TR. make_random_timing.py -num_stim 3 -num_runs 4 -run_time 200 \\ -stim_dur 2.0 -num_reps 8 -prefix stimesD \\ -pre_stim_rest 20 -post_stim_rest 20 -tr_locked -tr 2.0 5. Esoteric example. Similar to #2, but require an additional 0.7 seconds of rest after each stimulus (exactly the same as adding 0.7 to the stim_dur), set the granularity of random sequencing to 0.001 seconds, apply a random number seed of 31415, and set the verbose level to 2. Save a 3dDeconvolve -nodata command in @cmd.3dd . make_random_timing.py -num_stim 3 -num_runs 4 -run_time 200 \\ -stim_dur 3.5 -num_reps 8 -prefix stimesE \\ -pre_stim_rest 20 -post_stim_rest 20 \\ -min_rest 0.7 -max_rest 7.0 \\ -t_gran 0.001 -seed 31415 -verb 2 \\ -show_timing_stats -save_3dd_cmd @cmd.3dd 6. Example with varying number of events, durations and run times. ** Note that this does not make for a balanced design. Similar to #2, but require each stimulus class to have a different number of events. Class #1 will have 8 reps per run, class #2 will have 10 reps per run and class #3 will have 15 reps per run. The -num_reps option takes either 1 or -num_stim parameters. Here, 3 are supplied. make_random_timing.py -num_stim 3 -num_runs 4 \\ -run_time 200 190 185 225 \\ -stim_dur 3.5 4.5 3 -num_reps 8 10 15 \\ -pre_stim_rest 20 -post_stim_rest 20 \\ -prefix stimesF 7. Catch trials. If every time a main stimulus 'M' is presented it must follow another stimulus 'C', catch trials can be used to separate them. If the TRs look like ...CM.CM.....CM...CMCM, it is hard to separate the response to M from the response to C. When separate C stimuli are also given, the problem becomes simple : C..CM.CM...C.CM...CMCM. Now C and M can be measured separately. In this example we have 4 8-second main classes (A1, A2, B1, B2) that always follow 2 types of 8-second catch classes (A and B). The times of A1 are always 8 seconds after the times for A, for example. Main stimuli are presented 5 times per run, and catch trials are given separately an additional 4 times per run. That means, for example, that stimulus A will occur 14 times per run (4 as 'catch', 5 preceding A1, 5 preceding A2). Each of 3 runs will last 9 minutes. Initially we will claim that A1..B2 each lasts 16 seconds. Then each of those events will be broken into a 'catch' event at the beginning, followed by a 'main' event after another 8 seconds. Set the minimum time between any 2 events to be 1.5 seconds. Do this in 4 steps: a. Generate stimulus timing for 6 classes: A, B, A1, A2, B1, B2. Stim lengths will be 8, 8, and 16, 16, 16, 16 seconds, at first. Note that both the stimulus durations and frequencies will vary. make_random_timing.py -num_stim 6 -num_runs 3 -run_time 540 \\ -stim_dur 8 8 16 16 16 16 -num_reps 4 4 5 5 5 5 \\ -stim_labels A B A1 A2 B1 B2 -min_rest 1.5 -seed 54321 \\ -prefix stimesG b. Separate 'catch' trials from main events. Catch trails for A will occur at the exact stim times of A1 and A2. Therefore all of our time for A/A1/A2 are actually times for A (and similarly for B). Concatenate the timing files and save them. 1dcat stimesG_??_A.1D stimesG_??_A?.1D > stimesG_A_all.1D 1dcat stimesG_??_B.1D stimesG_??_B?.1D > stimesG_B_all.1D Perhaps consider sorting the stimulus times per run, since the 1dcat command does not do that. Use timing_tool.py. The new 'sorted' timing files would replace the 'all' timing files. timing_tool.py -timing stimesG_A_all.1D -sort \\ -write_timing stimesG_A_sorted.1D timing_tool.py -timing stimesG_B_all.1D -sort \\ -write_timing stimesG_B_sorted.1D c. To get stim times for the 'main' regressors we need to add 8 seconds to every time. Otherwise, the times will be identical to those in stimesG.a_03_A?.1D (and B). There are many ways to add 8 to the timing files. In this case, just run the program again, with the same seed, but add an offset of 8 seconds to all times. Then simply ignore the new files for A and B, while keeping those of A1, A2, B1 and B2. Also, save the 3dDeconvolve command to run with -nodata. make_random_timing.py -num_stim 6 -num_runs 3 -run_time 540 \\ -stim_dur 8 8 16 16 16 16 -num_reps 4 4 5 5 5 5 \\ -stim_labels A B A1 A2 B1 B2 -min_rest 1.5 -seed 54321 \\ -offset 8.0 -save_3dd_cmd @cmd.3dd.G -prefix stimesG d. Finally, fix the 3dDeconvolve command in @cmd.3dd.G. 1. Use timing files stimesG_A_sorted.1D and stimesG_B_sorted.1D from step b, replacing stimesG_01_A.1D and stimesG_01_B.1D. 2. Update the stimulus durations of A1, A2, B1 and B2 from 16 seconds to the correct 8 seconds (the second half of the 16 second intervals). This is necessary because the command in step (c) does not know about the updated A/B files from step (b). The first half of each 16 second A1/A2 stimulus is actually stimulus A, while the second half is really A1 or A2. Similarly for B. The resulting files are kept (and applied in and 3dDeconvolve commands): stimesG_[AB]_sorted.1D : the (sorted) 'catch' regressors, 14 stimuli per run (from step b) stimesG_*_[AB][12].1D : the 4 main regressors (at 8 sec offsets) (from step c) --- end of (long) example #7 --- 8. Example requiring partially fixed stimulus ordering. Suppose we have 2 sets of stimuli, question/answer/score along with face/doughnut. Anytime a question is given it is followed by an answer (after random rest) and then a score (after random rest). The face and doughnut stimuli are random, but cannot interrupt the q/a/s triples. Effectively, this means question, face and doughnut are random, but answer and score must always follow question. Rest should be randomly distributed anywhere. The q/a/s stimuli are each 1.5 seconds, but since we require a minimum of 1 second after 'q' and 'a', and 1.5 seconds after 's', those stimulus durations are given as 2.5, 2.5 and 3.0 seconds, respectively. The 'f' and 'd' stimuli are each 1 second. Each stimulus has 8 repetitions per run, over 4 240 second runs. The first and last 20 seconds of each run will be left to rest. make_random_timing.py -num_runs 4 -run_time 240 \\ -num_stim 5 -num_reps 8 \\ -stim_labels question answer score face doughnut \\ -stim_dur 2.5 2.5 3 1 1 \\ -ordered_stimuli question answer score \\ -pre_stim_rest 20 -post_stim_rest 20 \\ -show_timing_stats -seed 31415 -prefix stimesH To verify the stimulus order, consider using timing_tool.py to convert timing files to an event list. The corresponding command might be the following, output on a TR grid of 1.0 s. timing_tool.py -multi_timing stimesH*.1D \\ -multi_timing_to_events events.stimesH.txt \\ -multi_stim_dur 2.5 2.5 3 1 1 \\ -tr 1.0 -min_frac 0.5 -per_run -run_len 240 9. TR-locked example, fixed seed, limited consecutive events. Similar to #4, but restrict the number of consecutive events of each type to 2. make_random_timing.py -num_stim 3 -num_runs 2 -run_time 200 \\ -stim_dur 2.0 -num_reps 10 30 10 -prefix stimesI \\ -pre_stim_rest 20 -post_stim_rest 20 -tr_locked -tr 2.0 \\ -max_consec 2 ---------------------------------------------------------------------- NOTE: distribution of ISI To picture the distribution, consider the probability of starting with r rest events, given R total rest events and T total task events. The probability of starting with 0 rest events is actually the maximum, and equals the probability of selecting a task event first, which is T/(T+R). Let X be a random variable indicating the number of rest events to start a run. Then P(X=0) = T/(T+R). While this may look "large" (as in possibly close to 1), note that typically R >> T. For example, maybe there are 50 task events and 1000 rest "events" (e.g. 0.1 s, each). Then P(X=0) = 50/1050 = 0.0476. This ratio is generally closer to T/R than to 1.0. T/R is 0.05 here. More details... To take one step back, viewing this as the probability of having t task events among the first n events, it follows a hypergeometric distribution. That is because for each event type that is selected, there are fewer such events of that type remaining for subsequent selections. The selection is done *without* replacement. The total numbers of each type of class are fixed, as is the total rest. This differs it from the binomial distribution, where selection is done *with* replacement. Taking a simplistic view, go back to the probability of starting with exactly r rest events, as stated at the beginning. That means starting with r rest events followed by one task event. That means first choosing r rest events ((R choose r) / ((R+T) choose r)), then choosing one task event, T/(R+T-r). (R) (r) T R! (R+T-r-1)! P(X=r) = ----- * ------ = ----- * T * ---------- (R+T) (R+T-r) (R-r)! (R+T)! (r ) While this may not provide much insight on its own, consider the ratio of incremental probabilities P(X=r+1) / P(X=r): P(X=r+1) R-r R - r -------- = ------- = for visual significance = ----------- P(X=r) R+T-1-r R+T-1 - r The left side of that ratio is fixed at R/(R+T-1) = 1000/(1049) = .953 for the earlier example. It may by common to be in that ballpark. For subsequent r values, that ratio goes down, eventually hitting 0 when the rest is exhausted (r=R). This means that the distribution of such rest actually falls _below_ an exponential decay curve. It is close to (R/(R+T-1))^r at first, decaying more rapidly until hitting 0. ==> The overall distribution of ISI rest looks like an exponential decay curve, with a peak at r=0 (no rest) and probability close to T/R. Note that the average ISI should be approximately equal to total rest time / # task events (e.g. 100s / 50 stimuli = 2s (per stim)). So the cumulative distribution function would hit 0.5 where r corresponds to this ratio, e.g. r = 20, where each rest event is 0.1s. Test this: Create a histogram of all ISI durations based on 100 2-second events in a single run of length 300 (so 200 s for task, 100 s for rest), with rest distributed randomly on a 0.1 s time grid. Note that what matters is the number of stim events (100) and the number of rest events (1000), not their respective durations (unless there are user-imposed limits). Given the timing, "timing_tool.py -multi_timing_to_event_list" can be used to output ISIs (for example). Use that to simply make a list of ISIs, and then make a histogram. Let us repeat the process of generating events and ISIs, accumulating a list of ISIs, a total of 100 times. The generate and plot of histogram of all ISI duration counts. Since rest is on a 0.1 s grid, we will scale by 10 and make an integer histogram. echo -n "" > isis_all.1D foreach rep ( `count 1 100` ) echo simulation $rep make_random_timing.py -num_stim 1 -num_runs 1 -run_time 300 \\ -stim_dur 2 -num_reps 100 -prefix t -verb 0 ( timing_tool.py -multi_timing t_01.1D -multi_stim_dur 2 \\ -multi_timing_to_event_list GE:o - -verb 0 \\ | 1deval -a - -expr '10*a' >> isis_all.1D ) >& /dev/null end 3dhistog -int isis_all.1D | tee isis_hist.1D 1dplot -sepscl isis_hist.1D'[1,2]' Note that the histogram might be scaled down by a factor of 100 to get an expected ISI frequency per run (since we effectively accumulated the ISI lists over 100 runs). Basically, we are looking for something like a exponential decay curve in the frequency histogram (the lower plot). Include plot of probabilities, computed incrementally (no factorials). Use the same event counts, 100 task and 1000 rest events. Truncate this histogram to plot them together. set nhist = `1dcat isis_hist.1D | wc -l` make_random_timing.py -verb 0 -show_isi_pdf 100 1000 > pure_probs.1D grep -v prob pure_probs.1D | grep -v result | grep -v '\-----' \\ | head -n $nhist > prob.1D 1dplot -sepscl prob.1D'[1]' isis_hist.1D'[1,2]' Side note assuming replacement and the binomial distribution: In the case of replacement, we get a binomial distribution. In the same P(X=r) case (starting with r rest events), the probabilities are simple. P(X=r) = [R/(R+T)]^r * T/(R+T) Each rest probability is simply R/(R+T), while task is T/(R+T). The incremental probability is simply that of getting one more rest, which is R/(R+T) because of independence (samples are "replaced"). In this case, the PDF should more exactly follow an exponential decay curve. ---------------------------------------------------------------------- informational arguments: -help : display this help -help_advanced : display help for advanced usage -help_todo : display list of things to do -hist : display the modification history -show_valid_opts : display all valid options (short format) -ver : display the version number ---------------------------------------- advanced arguments/options: -help_advanced : display help for advanced usage -help_todo : "to do" list is mostly for advanced things -add_timing_class : create a new timing class (stim or rest) -add_stim_class : describe a new stimulus class (timing, etc.) -rand_post_stim_rest yes/no : allow rest after final stimulus -show_rest_events : show details of rest timing, per type -write_event_list FILE : create FILE listing all events and times -save_3dd_cmd FILE : write 3dDeconvolve script to FILE -make_3dd_contrasts : include pairwise contrasts in 3dD script ---------------------------------------- required arguments: -num_runs NRUNS : set the number of runs e.g. -num_runs 4 Use this option to specify the total number of runs. Output timing files will have one row per run (for -local_times in 3dDeconvolve). -run_time TIME : set the total time, per run (in seconds) e.g. -run_time 180 e.g. -run_time 180 150 150 180 This option specifies the total amount of time per run, in seconds. This time includes all rest and stimulation. This time is per run, even if -across_runs is used. -num_stim NSTIM : set the number of stimulus classes e.g. -num_stim 3 This specifies the number of stimulus classes. The program will create one output file per stimulus class. -num_reps REPS : set the number of repetitions (per class?) e.g. -num_reps 8 e.g. -num_reps 8 15 6 This specifies the number of repetitions of each stimulus type, per run (unless -across_runs is used). If one parameter is provided, every stimulus class will be given that number of repetitions per run (unless -across_runs is given, in which case each stimulus class will be given a total of that number of repetitions, across all runs). The user can also specify the number of repetitions for each of the stimulus classes separately, as a list. see also: -across_runs -prefix PREFIX : set the prefix for output filenames e.g. -prefix stim_times --> might create: stim_times_001.1D The option specifies the prefix for all output stimulus timing files. The files will have the form: PREFIX_INDEX[_LABEL].1D, where PREFIX is via this option, INDEX is 01, 02, ... through the number of stim classes, and LABEL is optionally provided via -stim_labels. Therefore, output files will be sorted alphabetically, regardless of any labels, in the order that they are given to this program. see also -stim_labels -show_timing_stats : show statistics from the timing e.g. -show_timing_stats If this option is set, the program will output statistical information regarding the stimulus timing, and on ISIs (inter-stimulus intervals) in particular. One might want to be able to state what the min, mean, max and stdev of the ISI are. -stim_dur TIME : set the duration for a single stimulus e.g. -stim_dur 3.5 e.g. -stim_dur 3.5 1.0 4.2 This specifies the length of time taken for a single stimulus, in seconds. These stimulation intervals never overlap (with either rest or other stimulus intervals) in the output timing files. If a single TIME parameter is given, it applies to all of the stimulus classes. Otherwise, the user can provide a list of durations, one per stimulus class. ---------------------------------------- optional arguments: -across_runs : distribute stimuli across all runs at once e.g. -across_runs By default, each of -num_stim stimuli are randomly distributed within each run separately, per class. But with the -across_runs option, these stimuli are distributed across all runs at once (so the number of repetitions per run will vary). For example, using -num_stim 2, -num_reps 24 and -num_runs 3, assuming -across_runs is _not_used, there would be 24 repetitions of each stim class per run (for a total of 72 repetitions over 3 runs). However, if -across_runs is applied, then there will be only the 24 repetitions over 3 runs, for an average of 8 per run (though there will probably not be exactly 8 in every run). -make_3dd_contrasts : add all pairwise contrasts to 3dDeconvolve This option is particularly useful if make_random_timing.py is part of an experiment design search script. In any case, this option can be used to add all possible pairwise contrasts to the 3dDeonvolve command specified by -save_3dd_cmd. Options -save_3dd_cmd and -stim_labels are also required. -max_consec c1 c2 ... cn : specify maximum consecutive stimuli per class e.g. A. -max_consec 2 e.g. B. -max_consec 2 2 2 2 e.g. C. -max_consec 0 4 2 0 This option is used to limit the number of consecutive events of one or more classes. Assuming 4 stimulus classes, examples A and B limit each event type to having at most 2 consecutive events of that type. Example C shows limiting only the second and third stimulus classes to consecutive events of length 4 and 2, respectively. A limit of 0 means no limit (num_reps, effectively). -max_rest REST_TIME : specify maximum rest between stimuli e.g. -max_rest 7.25 This option applies a second phase in ordering events. After events have been randomized, non-pre- and non-post-stim rest periods are limited to the max_rest duration. Any rest intervals exceeding this duration are distributed randomly into intervals below this maximum. -min_rest REST_TIME : specify extra rest after each stimulus e.g. -min_rest 0.320 --> would add 320 milliseconds of rest after each stimulus There is no difference between applying this option and instead adding the REST_TIME to that of each regressor. It is merely another way to partition the stimulus time period. For example, if each stimulus lasts 1.5 seconds, but it is required that at least 0.5 seconds separates each stimulus pair, then there are 2 equivalent ways to express this: A: -stim_dur 2.0 B: -stim_dur 1.5 -min_rest 0.5 These have the same effect, but perhaps the user wants to keep the terms logically separate. However the program simply adds min_rest to each stimulus length. -offset OFFSET : specify an offset to add to every stim time e.g. -offset 4.5 Use this option to offset every stimulus time by OFFSET seconds. -ordered_stimuli STIM1 STIM2 ... : specify a partial ordering of stimuli e.g. -ordered_stimuli primer choice reward e.g. -ordered_stimuli 4 2 5 e.g. -ordered_stimuli stimA replyA -ordered stimuli stimB replyB e.g. -ordered_stimuli 1 2 -ordered_stimuli 3 4 -ordered_stimuli 5 6 This option is used to require that some regressors are ordered. For example, every time a question stimulus occurs it is followed by a response stimulus, with only random rest in between. There might be other stimuli, but they cannot break the question/response pair. So all the stimuli and rest periods are still random, except that given regressors must maintain the specified order. Given the first example, whenever primer occurs it is followed first by choice and then by reward. Other stimuli might come before primer or after reward, but not in between. In the third example the stim/reply pairs are never broken, so stimA and replyA are always together, as are stimB and replyB. Note: - Multiple -ordered_stimuli options may be used. - A single stimulus may not appear in more than one such option. - Stimulus entries can be either labels (requiring -labels to be specified first) or 1-based indices, running from 1..N. See example 8 above. -pre_stim_rest REST_TIME : specify minimum rest period to start each run e.g. -pre_stim_rest 20 Use this option to specify the amount of time that should pass at the beginning of each run before the first stimulus might occur. The random placing of stimuli and rest will occur after this time in each run. As usual, the time is in seconds. -post_stim_rest REST_TIME : specify minimum rest period to end each run e.g. -post_stim_rest 20 Use this option to specify the amount of time that should pass at the end of each run after the last stimulus might occur. One could consider using -post_stim_rest of 12.0, always, to account for the decay of the BOLD response after the last stimulus period ends. Note that the program does just prevent a stimulus from starting after this time, but the entire stimulation period (described by -stim_dur) will end before this post_stim_rest period begins. For example, if the user provides "-run_time 100", "-stim_dur 2.5" and "-post_stim_rest 15", then the latest a stimulus could possibly occur at is 82.5 seconds into a run. This would allow 2.5 seconds for the stimulus, plus another 15 seconds for the post_stim_rest period. -save_3dd_cmd FILENAME : save a 3dDeconvolve -nodata example e.g. -save_3dd_cmd sample.3dd.command Use this option to save an example of running "3dDeconvolve -nodata" with the newly created stim_times files. The saved script includes creation of a SUM regressor (if more than one stimulus was given) and a suggestion of how to run 1dplot to view the regressors created from the timing files. The use of the SUM regressor is to get a feel for what the expected response might look at a voxel that response to all stimulus classes. If, for example, the SUM never goes to zero in the middle of a run, one might wonder whether it is possible to accurately separate each stimulus response from the baseline. -seed SEED : specify a seed for random number generation e.g. -seed 3141592 This option allows the user to specify a seed for random number generation in the program. The main reason to do so is to be able to duplicate results. By default, the seed is based on the current system time. -stim_labels LAB1 LAB2 ... : specify labels for the stimulus classes e.g. -stim_labels houses faces donuts Via this option, one can specify labels to become part of the output filenames. If the above example were used, along with -prefix stim, the first stimulus timing would be written to stim_01_houses.1D. The stimulus index (1-based) is always part of the filename, as that keeps the files alphabetical in the order that the stimuli were specified to the program. There must be exactly -num_stim labels provided. -t_digits DIGITS : set the number of decimal places for times e.g. -t_digits 3 e.g. -t_digits -1 Via this option one can control the number of places after the decimal that are used when writing the stimulus times to each output file. The special value of -1 implies %g format. The default is 1, printing times in tenths of a second. But if a higher time granularity is requested via -t_gran, one might want more places after the decimal. Note that if a user-supplied -t_gran does not round to a tenth of a second, the default t_digits changes to 3, to be in milliseconds. -t_gran GRANULARITY : set the time granularity e.g. -t_gran 0.001 The default time granularity is 0.1 seconds, and rest timing is computed at that resolution. This option can be applied to change the resolution. There are good reasons to go either up or down. One might want to use 0.001 to obtain a temporal granularity of a millisecond, as times are often given at that resolution. Also, one might want to use the actual TR, such as 2.5 seconds, to ensure that rest and stimuli occur on the TR grid. Note that such a use also requires -stim_dur to be a multiple of the TR. -tr TR : set the scanner TR e.g. -tr 2.5 The TR is needed for the -tr_locked option (so that all times are multiples of the TR), and for the -save_3dd_cmd option (the TR must be given to 3dDeconvolve). see also: -save_3dd_cmd, -tr_locked -verb LEVEL : set the verbose level e.g. -verb 2 The default level is 1, and 0 is consider 'quiet' mode, only reporting errors. The maximum level is currently 4. - R Reynolds May 7, 2008 motivated by Ikuko Mukai =========================================================================== """ g_help_advanced = """ =========================================================================== make_random_timing.py - Advanced usage With advanced usage, timing classes are defined for both stimulus periods and rest periods. Timing classes specify duration types that have different distributions (min, mean, max and distribution type), which can be applied to stimulus events or to rest events. When specifying a timing class, one can provide: min : min, mean and maximum for possible durations mean (note that for a uniform distribution, the mean or max max implies the other, while that is not true for decay) dtype : distribution type (default=decay) decay: shorter events are more likely see "NOTE: distribution of ISI" * new method as of Feb 3, 2017 decay_old: old decay method, which can bunch up at max limit, if one is applied uniform_rand: randomly chosen durations with uniform dist uniform_grid: durations spread evenly across grid fixed: one duration is specified INSTANT: duration = 0 t_grid : all durations are fixed on this grid (default=0.01s) One can provide subsets: min : implies fixed min, mean max : implies decay on default t_grid min, mean max, dtype : implies default t_trid min, mean max, dtype, t_grid Every stimulus class type is followed by a fixed rest class type. So rest periods are "attached" to the preceding stimulus periods. For example, the 'faces' class events might last for 0.5 - 1.5 seconds, and be uniformly distributed (so average = 1s). Those face events might then be followed by 0.5 - 8.5 seconds of rest with a 'decay' distribution (so shorter durations are more probable than longer durations). The 'decay' distribution type matches that of the basic (non-advanced) use this program. See "NOTE: distribution of ISI" in the -help output. ------------------------------------------------------- Advanced Example 1: basic, with 3 conditions - This is a simple case with 3 conditions, each having 8 events per run of duration 3.5 s. Rest is randomly distributed using the default 'decay' distribution (meaning shorter periods are more likely than longer ones). The first and last 20 s is also allocated for rest. - Do this for 4 runs of length 200 s each. - Also, do not allow any extra rest (beyond the specified 10 s) after the final stimulus event. - Generate 3dDeconvolve command script (and with pairwise contrasts). - Show timing statistics. Save a complete event list (events.adv.1.txt). make_random_timing.py -num_runs 4 -run_time 200 \\ -pre_stim_rest 10 -post_stim_rest 10 \\ -rand_post_stim_rest no \\ -add_timing_class stim 3.5 \\ -add_timing_class rest 0 -1 -1 \\ -add_stim_class houses 10 stim rest \\ -add_stim_class faces 10 stim rest \\ -add_stim_class donuts 10 stim rest \\ -show_timing_stats \\ -write_event_list events.adv.1.txt \\ -save_3dd_cmd cmd.3dd.eg1.txt \\ -make_3dd_contrasts \\ -seed 31415 -prefix stimes.adv.1 ------------------------------------------------------- Advanced Example 2: varying stimulus and rest timing classes - This has 4 stimulus conditions employing 3 different stimulus timing classes and 3 different rest timing classes. timing classes (stim and rest periods): stima: durations in [0.5, 10], ave = 3s (decay distribution) stimb: durations in [0.1, 3], ave = 0.5s (decay distribution) stimc: durations of 2s resta: durations in [0.2, 1.2], ave = 0.7 (uniform rand dist) restb: durations in [0.5, 1.5], ave = 1.0 (uniform grid dist) restc: durations in (0, inf) (decay dist) - absorbs remaining rest conditions (each has stim timing type and subsequent rest timing type) # events (per run) stim timing rest timing -------- ----------- ----------- houses : 20 stima resta faces : 20 stimb restb donuts : 20 stimb restb pizza : 20 stimc restc - Do not allow any rest (aside from -post_stim_rest) after final stim (per run). So there will be exactly the rest from -post_stim_rest at the end of each run, 10s in this example. make_random_timing.py -num_runs 2 -run_time 300 \\ -pre_stim_rest 10 -post_stim_rest 10 \\ -rand_post_stim_rest no \\ -add_timing_class stima 0.5 3 10 \\ -add_timing_class stimb 0.1 0.5 3 \\ -add_timing_class stimc 2 \\ -add_timing_class resta 0.2 .7 1.2 dist=uniform_rand \\ -add_timing_class restb 0.5 1 1.5 dist=uniform_grid \\ -add_timing_class restc 0 -1 -1 \\ -add_stim_class houses 20 stima resta \\ -add_stim_class faces 20 stimb restb \\ -add_stim_class donuts 20 stimb restb \\ -add_stim_class pizza 20 stimc restc \\ -write_event_list events.adv.2 \\ -show_timing_stats \\ -seed 31415 -prefix stimes.adv.2 ------------------------------------------------------- Advanced Example 3: ordered event types - Every cue event is followed by test and then result. - Every pizza1 event is followed by pizza2 and then pizza3. - The stimc timing class has durations on a grid of 0.1s, rather than the default of 0.01s. - Write a corresponding 3dDeconvolve script, cmd.3dd.eg3.txt. make_random_timing.py -num_runs 2 -run_time 300 \\ -pre_stim_rest 10 -post_stim_rest 10 \\ -rand_post_stim_rest no \\ -add_timing_class stima 0.5 3 10 \\ -add_timing_class stimb 0.1 0.5 3 \\ -add_timing_class stimc 0.1 2.5 10 t_gran=0.1 \\ -add_timing_class stimd 2 \\ -add_timing_class resta 0.2 .7 1.2 dist=uniform_rand \\ -add_timing_class restb 0.5 1 1.5 dist=uniform_grid \\ -add_timing_class restc 0 -1 -1 \\ -add_stim_class cue 20 stima resta \\ -add_stim_class test 20 stimb restb \\ -add_stim_class result 20 stimb restb \\ -add_stim_class pizza1 10 stimc restc \\ -add_stim_class pizza2 10 stimc restc \\ -add_stim_class pizza3 10 stimc restc \\ -add_stim_class salad 10 stimd restc \\ -write_event_list events.adv.3 \\ -show_timing_stats \\ -ordered_stimuli cue test result \\ -ordered_stimuli pizza1 pizza2 pizza3 \\ -save_3dd_cmd cmd.3dd.eg3.txt \\ -seed 31415 -prefix stimes.adv.3 ------------------------------------------------------- Advanced Example 4: limit consecutive events per class type - Use simple 1s stim events and random rest (decay). - For entertainment, houses/faces and tuna/fish are ordered event pairs. - Classes houses, faces, tuna and fish are restricted to a limit of 3 consecutive events. - There is no limit on donuts. Why would there be? make_random_timing.py -num_runs 2 -run_time 600 \\ -pre_stim_rest 0 -post_stim_rest 0 \\ -add_timing_class stim 1 \\ -add_timing_class rest 0 -1 -1 \\ -add_stim_class houses 100 stim rest \\ -add_stim_class faces 100 stim rest \\ -add_stim_class tuna 100 stim rest \\ -add_stim_class fish 100 stim rest \\ -add_stim_class donuts 100 stim rest \\ -ordered_stimuli houses faces \\ -ordered_stimuli tuna fish \\ -max_consec 3 3 3 3 0 \\ -show_timing_stats \\ -write_event_list events.adv.4 \\ -seed 31415 -prefix stimes.adv.4 -verb 2 --------------------------------------------------------------------- options (specific to the advanced usage): -help_advanced : display help for advanced usage -help_todo : "to do" list is mostly for advanced things -add_timing_class : create a new timing class (stim or rest) -add_stim_class : describe a new stimulus class (timing, etc.) -rand_post_stim_rest yes/no : allow rest after final stimulus -show_rest_events : show details of rest timing, per type -write_event_list FILE : create FILE listing all events and times ---------------------------------------------------------------------- R Reynolds Jan 20, 2017 motivated by K Kircanski and A Stringaris =========================================================================== """ g_history = """ make_random_timing.py history: 0.1 May 07, 2008: initial release 0.2 May 18, 2008: - changed -stim_time option to -stim_dur - added options -tr_locked, -tr and -save_3dd_cmd 0.3 June 6, 2008: get_*_opt now returns error code 0.4 Oct 27, 2008: - actually implemented -min_rest, which I apparently forgot to do - added -offsets option 0.5 Oct 31, 2008: - added -show_timing_stats option - made small change that affects timing (old results will not match) 0.6 Dec 01, 2008: - moved min_mean_max_stdev to afni_util.py - modified examples to correspond with those in timing_tool.py 0.7 Dec 24, 2008: redefine 'sum' for old python versions 0.8 Feb 05, 2009: added timing_tool.py use to catch trial example 0.9 Apr 17, 2009: added -ordered_stimuli 0.10 Jun 17, 2009: - print block durations with general decimal accuracy - added -make_3dd_contrasts 1.0 Jul 14, 2009: call this a release version - added -max_rest for Steffen S (see message board post 30189) 1.1 May 01, 2010: added -max_consec for Liat of Cornell 1.2 Aug 24, 2010: small updates for 3dDeconvolve -nodata script - update polort and write -nodata TR using 3 decimal places 1.3 Mar 09, 2011: fixed bug writing comment text in 3dD script - noted by Z Saad and P Kaskan 1.4 Jun 08, 2011: tr-lock test: fixed print and added min_rest to durs 1.5 Apr 08, 2012: -ordered_stimuli can now use labels 1.6 May 01, 2012: -ordered_stimuli now works with -max_consec - requested by Liat 1.7 Oct 03, 2012: some options do not allow dashed parameters 1.8 Nov 14, 2012: fixed checks for random space in -max_consec case 1.9 Aug 21, 2015: added help for understanding the distribution of ISI see: NOTE: distribution of ISI 1.10 Jun 01, 2016: minor updates to verbose output 2.0 Jan 20, 2017: basically a new program * Advanced usage: applying user-defined timing classes for stim/rest. see: make_random_timing.py -help_advanced 2.1 Jan 23, 2017: allow use of INSTANT timing class; reorder example opts 2.2 Feb 3, 2017: - decay class now follows a better curve - added decay_old class for old decay method 2.3 May 9, 2017: applied -offset for advanced case 2.4 May 24, 2017: advanced -save_3dd_cmd and -make_3dd_contrasts """ g_version = "version 2.4, May 24, 2017" g_todo = """ - add -show_consec_stats option? - c23 shows small post-stim rest... - reconcile t_grid as global vs per class (init/pass as single parameters) - make new method for decay that better handles max duration, w/out spike - add warning if post-stim rest < 3 seconds - add option to specify timing classes for pre and post stim rest - add related dist_types rand_unif and rand_gauss? - global "-across_runs" should still apply - maybe this program should just be a new one? gen_random_timing.py? - MRT could be forked and depricated, but maybe people want to use it? - or perhaps it could be clear by calling it old_MRT.py -no, just try to partition options - OLD: -num_stim, -num_reps, -stim_dur, -stim_labels, -min_rest - ADV: -add_stim_class, -add_timing_class """ gDEF_VERB = 1 # default verbose level gDEF_OLD_T_GRAN = 0.1 # old-style time granularity, in seconds gDEF_OLD_DEC_PLACES = 1 # old-style digits after decimal for times gDEF_MIN_T_GRAN = 0.0001 # minimum time granularity, in seconds g_instant_timing_class = LRT.TimingClass('INSTANT', 0, 0, 0) g_style_opts_old = ['-num_stim', '-num_reps', '-stim_dur', '-stim_labels', '-min_rest'] g_style_opts_new = ['-add_stim_class', '-add_timing_class'] # sample usage: # # -add_timing_class stimA 3 # -add_timing_class stimA 3 3 3 decay 0.1 # -add_timing_class stimA 3 5 10 # -add_timing_class stimA 3 5 10 decay # -add_timing_class stimA 3 5 10 decay 0.1 # -add_timing_class stimA 3 5 7 uniform 1 # # -add_timing_class restA 3 # -add_timing_class restA 3 5 10 decay # -add_timing_class restA 1 5 9 uniform 0.1 (decide how to sample) # order of stim evens should be controlled by: # - randomization # - max consec # - ordered_stimuli # - across_runs (sum (sdur+ave rest) to match run lengths) # - catch trials (can do later) # # stim events: # randomize events subject to any constraints (max consec, ordered stim) # compute random timing for events # apply rest: # add all minimum rest: is design possible? # add all maximum rest: is design possible? # class RandTiming: def __init__(self, label): # actual stimulus timing lists self.stimes = [] self.fnames = [] # output filenames # general parameters self.label = label self.verb = gDEF_VERB self.valid_opts = None # OptionList self.user_opts = None # ------------------------------------------------------------ # required arguments for advanced method self.tclasses = [g_instant_timing_class] # TimingClass instances self.sclasses = [] # StimClass instances self.advanced = 0 # advanced options self.rand_post_stim_rest = 1 # include random rest from last event? self.show_rest_events = 0 # show stats and rest events # pre- and post-rest timing classes self.pre_stimc = g_instant_timing_class self.post_stimc = g_instant_timing_class # other parameters for advanced method self.stim_event_list = [] # stim index, duration lists self.full_event_list = [] # index, duration, time lists self.stim_adata = [] # AfniData instance, per stim class self.stimdict = {} # name:index Stim dict, for name lookup # ------------------------------------------------------------ # required arguments for basic method self.num_reps = [] # number of stimuli, per class (per run) self.stim_dur = [] # time of single stimulus (seconds) # - per stim class # ------------------------------------------------------------ # other required arguments self.num_stim = 0 # number of stimulus classes self.num_runs = 0 # number of runs self.prefix = None # prefix for output files # add .LABEL.INDEX.1D self.run_time = [] # total time per run (seconds) # ------------------------------------------------------------ # optional arguments self.across_runs = 0 # flag: stimuli span all runs self.pre_stim_rest = 0 # seconds before first stim self.post_stim_rest = 0 # seconds after last stim self.max_rest = 0.0 # maximum rest after each stimulus self.min_rest = 0.0 # minimum rest after each stimulus self.offset = 0.0 # offset for all stimulus times self.seed = None # random number seed self.orderstim= [] # list of ordered stimulus lists (ints) self.orderlabs= [] # list of ordered stimulus lists (strs) self.osleaders= [] # leaders and followers self.osfollow = [] self.osdict = {} # list of followers per leader self.max_consec = [] # max consectutive stimuli per type self.t_gran = gDEF_OLD_T_GRAN # time granularity for rest self.tgset = 0 # was this field set? self.t_digits = gDEF_OLD_DEC_PLACES # digits after decimal for times # (-1 means to use %g) self.labels = None # labels to be applied to filenames self.tr_locked = 0 # flag: require TR-locked timing self.tr = 0.0 # TR, for use in output 3dDecon cmd self.control_breaks = 1 # flag: across runs, add max stim time # to post_stim_rest self.file_3dd_cmd = None # file for 3dD -nodata command self.make_3dd_contr = 0 # flag to make all 3dD contrasts self.file_elist = '' # file for stimulus event list # statistics self.show_timing_stats = 0 # do we show ISI statistics? self.isi = [] # list of isi's, per run self.prerest = [] # pre-stim rest, per run self.postrest = [] # post-stim rest, per run def init_opts(self): global g_help_string self.valid_opts = OL.OptionList('for input') # short, terminal arguments self.valid_opts.add_opt('-help', 0, [], \ helpstr='display program help') self.valid_opts.add_opt('-help_advanced', 0, [], \ helpstr='display program help for ADVANCED usage') self.valid_opts.add_opt('-help_todo', 0, [], \ helpstr='display stupid todo list') self.valid_opts.add_opt('-hist', 0, [], \ helpstr='display the modification history') self.valid_opts.add_opt('-show_isi_pdf', 2, [], \ helpstr='show init ISI pdf given NTASK NREST') self.valid_opts.add_opt('-show_isi_f_pdf', 2, [], \ helpstr='show init fact ISI pdf given NTASK NREST') self.valid_opts.add_opt('-show_valid_opts', 0, [], \ helpstr='display all valid options') self.valid_opts.add_opt('-ver', 0, [], \ helpstr='display the current version number') # new 'required' arguments self.valid_opts.add_opt('-add_timing_class', -2, [], req=0, helpstr='timing class: name min mean max pdf tgran') self.valid_opts.add_opt('-add_stim_class', 4, [], req=0, helpstr='stim class: name nreps stiming rtiming') # new optional arguments self.valid_opts.add_opt('-rand_post_stim_rest', 1, [], req=0, acplist=['no','yes'], helpstr='include random rest after final stimulus? y/n') self.valid_opts.add_opt('-show_rest_events', 0, [], req=0, helpstr='show stats on each rest class (y/n)') self.valid_opts.add_opt('-write_event_list', 1, [], helpstr="write event list to file ('-' for stdout)") # old 'required' arguments self.valid_opts.add_opt('-num_stim', 1, [], req=0, helpstr='number of stimulus types') self.valid_opts.add_opt('-num_runs', 1, [], req=1, helpstr='number of scanning runs') self.valid_opts.add_opt('-num_reps', -1, [], req=0, okdash=0, helpstr='number of stimulus reps per run, per class') self.valid_opts.add_opt('-stim_dur', -1, [], req=0, okdash=0, helpstr='length of each stimulus, in seconds') # required arguments self.valid_opts.add_opt('-prefix', 1, [], req=1, helpstr='prefix for output stimulus timing files') self.valid_opts.add_opt('-run_time', -1, [], req=1, okdash=0, helpstr='total length of each run, in seconds') # optional arguments self.valid_opts.add_opt('-across_runs', 0, [], helpstr='distribute stim reps across all runs') self.valid_opts.add_opt('-make_3dd_contrasts', 0, [], helpstr='add contrasts pairs to 3dDeconvolve script') self.valid_opts.add_opt('-max_consec', -1, [], okdash=0, helpstr='max consecutive occurances of each stim type') self.valid_opts.add_opt('-max_rest', 1, [], helpstr='maximum rest time after each stimulus') self.valid_opts.add_opt('-min_rest', 1, [], helpstr='minimum rest time after each stimulus') self.valid_opts.add_opt('-offset', 1, [], helpstr='offset to add to every stimulus time') self.valid_opts.add_opt('-ordered_stimuli', -1, [], okdash=0, helpstr='require these stimuli to be so ordered') self.valid_opts.add_opt('-pre_stim_rest', 1, [], helpstr='time before first stimulus, in seconds') self.valid_opts.add_opt('-post_stim_rest', 1, [], helpstr='time after last stimulus, in seconds') self.valid_opts.add_opt('-save_3dd_cmd', 1, [], helpstr='file for "3dDeconvolve -nodata" command') self.valid_opts.add_opt('-seed', 1, [], helpstr='seed for random number generation (integer)') self.valid_opts.add_opt('-show_timing_stats', 0, [], helpstr='show statistics for inter-stimulus intervals') self.valid_opts.add_opt('-stim_labels', -1, [], okdash=0, helpstr='specify stimulus labels for filenames') self.valid_opts.add_opt('-t_digits', 1, [], helpstr='digits after decimal, for printing times') self.valid_opts.add_opt('-t_gran', 1, [], helpstr='time_granularity for rest, in seconds') self.valid_opts.add_opt('-tr', 1, [], helpstr='specify TR for 3dDeconvolve command'); self.valid_opts.add_opt('-tr_locked', 0, [], helpstr='specify TR and enforce TR-locked timing'); self.valid_opts.add_opt('-verb', 1, [], helpstr='verbose level (0=quiet, 1=default, ...)') def read_opts(self): """check for terminal arguments, then read the user options""" # process any optlist_ options self.valid_opts.check_special_opts(sys.argv) # ------------------------------------------------------------ # check for terminal arguments # if argv has only the program name, or user requests help, show it if len(sys.argv) <= 1 or '-help' in sys.argv: print g_help_string return 0 if '-help_advanced' in sys.argv: print g_help_advanced return 0 if '-help_todo' in sys.argv: print g_todo return 0 # check for -ver and -hist, too if '-hist' in sys.argv: print g_history return 0 if '-ver' in sys.argv: print g_version return 0 # maybe the user wants to list valid options if '-show_valid_opts' in sys.argv: self.valid_opts.show('', 1) return 0 # ---------------------------------------- # terminal processing args that involve processing if '-show_isi_pdf' in sys.argv or '-show_isi_f_pdf' in sys.argv: return do_isi_pdfs(sys.argv) # ------------------------------------------------------------ # read all user options self.user_opts = OL.read_options(sys.argv, self.valid_opts) if not self.user_opts: return 1 # error condition return None # normal completion def process_opts(self): """apply each option, and turn list options with single params into complete lists (e.g. stim_dur is per stim class)""" # ---------------------------------------- # set verb first self.verb, err = self.user_opts.get_type_opt(int, '-verb') if self.verb == None: self.verb = gDEF_VERB elif err: return 1 # ----------------------------------------------------------------- # determine advanced or basic usage styles and get relevant args # ADVANCED: -add_stim_class, -add_timing_class if self.has_any_opts(g_style_opts_new): if self.has_any_opts(g_style_opts_old): print '** have both new- and old-styled options, use only one' print ' ADVAN: %s' % ', '.join(g_style_opts_new) print ' BASIC: %s' % ', '.join(g_style_opts_old) return 1 # note that we have the advanced style self.advanced = 1 # get timing classes first, required for stim classes olist = self.user_opts.find_all_opts('-add_timing_class') if len(olist) == 0: print '** ADV STYLE: missing option -add_timing_class' return 1 for opt in olist: if self.apply_opt_timing_class(opt): return 1 olist = self.user_opts.find_all_opts('-add_stim_class') if len(olist) == 0: print '** ADV STYLE: missing option -add_stim_class' return 1 for opt in olist: if self.apply_opt_stim_class(opt): return 1 # default to including post-stim rest if self.user_opts.have_no_opt('-rand_post_stim_rest'): self.rand_post_stim_rest = 0 if self.user_opts.find_opt('-show_rest_events'): self.show_rest_events = 1 # and set some of the old-style parameters self.num_stim = len(olist) self.labels = [scl.name for scl in self.sclasses] # OLD: -num_stim, -num_reps, -stim_dur, -stim_labels, -min_rest elif self.has_any_opts(g_style_opts_old): self.num_stim, err = self.user_opts.get_type_opt(int, '-num_stim') if self.num_stim == None or err: return 1 # set num_reps list of length num_runs self.num_reps, err = self.user_opts.get_type_list(int, '-num_reps', self.num_stim, 'num_stim') if self.num_reps == None or err: return 1 if self.verb > 1: print UTIL.int_list_string(self.num_reps, '-- num_reps : ') # set stim_dur list of length num_stim self.stim_dur, err = self.user_opts.get_type_list(float, '-stim_dur', self.num_stim, 'stim_dur') if self.stim_dur == None or err: return 1 if self.verb > 1: print UTIL.gen_float_list_string(self.stim_dur, '-- stim_dur : ') # get any labels self.labels, err = self.user_opts.get_string_list('-stim_labels') if self.labels and len(self.labels) != self.num_stim: print '** error: %d stim classes but %d labels: %s' \ % (self.num_stim, len(self.labels), self.labels) return 1 self.min_rest, err = self.user_opts.get_type_opt(float,'-min_rest') if self.min_rest == None: self.min_rest = 0 elif err: return 1 # end: process OLD and ADV style options # ----------------------------------------------------------------- # ---------------------------------------- # required args - single parameter self.num_runs, err = self.user_opts.get_type_opt(int, '-num_runs') if self.num_runs == None or err: return 1 self.prefix, err = self.user_opts.get_string_opt('-prefix') if self.prefix == None or err: return 1 if self.verb > 1: print '-- have %d stim and %d runs, prefix = %s' % \ (self.num_stim, self.num_runs, self.prefix) # ---------------------------------------- # required args - (possibly) multiple parameter # set run_time list of length num_runs self.run_time, err = self.user_opts.get_type_list(float, '-run_time', self.num_runs, 'num_runs') if self.run_time == None or err: return 1 if self.verb > 1: print UTIL.gen_float_list_string(self.run_time, '-- run_time : ') # ---------------------------------------- # optional arguments (if failure, use default) # check for the 'across_runs' flag opt = self.user_opts.find_opt('-across_runs') if opt: self.across_runs = 1 opt = self.user_opts.find_opt('-make_3dd_contrasts') if opt: self.make_3dd_contr = 1 self.pre_stim_rest, err = self.user_opts.get_type_opt(float, '-pre_stim_rest') if self.pre_stim_rest == None: self.pre_stim_rest = 0.0 elif err: return 1 self.post_stim_rest, err = self.user_opts.get_type_opt(float, \ '-post_stim_rest') if self.post_stim_rest == None: self.post_stim_rest = 0.0 elif err: return 1 self.max_rest, err = self.user_opts.get_type_opt(float,'-max_rest') if self.max_rest == None: self.max_rest = 0 elif err: return 1 self.offset, err = self.user_opts.get_type_opt(float,'-offset') if self.offset == None: self.offset = 0 elif err: return 1 self.seed, err = self.user_opts.get_type_opt(float,'-seed') if self.seed == None: self.seed = None elif err: return 1 # maybe there is a limit on the number of consecutive events per type opt = self.user_opts.find_opt('-max_consec') if opt: self.max_consec,err = self.user_opts.get_type_list(int,'-max_consec', length=self.num_stim, len_name='num_stim', opt=opt) if err: return 1 if self.max_consec and self.verb > 1: print UTIL.int_list_string(self.max_consec, '-- max_consec : ') # gather a list of lists specified by -ordered_stimuli options oname = '-ordered_stimuli' olist = self.user_opts.find_all_opts(oname) for opt in olist: # check for non-int: if labels require labels to be already set slist, err = self.user_opts.get_type_list(int, opt=opt, verb=0) if err: # see if they are labels slist, err = self.user_opts.get_string_list(opt=opt) if err: print '** %s: need list of ints or labels' % oname return 1 # see if we have labels to convert into indices ilist, err = self.labels_to_indices(slist) if err: print '** %s requires indices or known labels' % oname print ' have: %s %s' % (oname, ' '.join(slist)) return 1 self.orderlabs.append(slist) slist = ilist self.orderstim.append(slist) if self.verb>1: print UTIL.int_list_string(slist, '-- orderstim (1-based): ') if self.verb > 1 or self.user_opts.find_opt('-show_timing_stats'): self.show_timing_stats = 1 self.t_gran, err = self.user_opts.get_type_opt(float,'-t_gran') if err: return 1 # check the result after -tr_locked self.tr, err = self.user_opts.get_type_opt(float,'-tr') if not self.tr: self.tr = 0.0 elif err: return 1 opt = self.user_opts.find_opt('-tr_locked') if opt: self.tr_locked = 1 if self.tr_locked: # set t_gran, and check stim_lengths if self.t_gran: print '** cannot use both -tr_locked and -t_gran' return 1 if self.tr == 0.0: print '** -tr_locked option requires -tr' return 1 self.t_gran = self.tr; if self.verb > 1: print '++ setting t_gran to TR, %0.1f s' % self.tr # fixed print and added min_rest to durations 8 Jun, 2011 sd = [dur + self.min_rest for dur in self.stim_dur] if not UTIL.vals_are_multiples(self.tr, sd, digits=4): print '** want TR-locked, but stim durations are not' \ ' multiples of TR %.3f' % self.tr print ' min_rest + duration(s): %s' % sd return 1 # if t_gran was set, note the fact, else apply default if self.t_gran: self.tgset = 1 else: self.t_gran = gDEF_OLD_T_GRAN # t_digits must come after t_gran is set self.t_digits, err = self.user_opts.get_type_opt(float,'-t_digits') if self.t_digits == None: if self.t_gran == round(self.t_gran,1): self.t_digits = gDEF_OLD_DEC_PLACES else: self.t_digits = 3 elif err: return 1 self.file_3dd_cmd, err = self.user_opts.get_string_opt('-save_3dd_cmd') if err: return 1 fname, err = self.user_opts.get_string_opt('-write_event_list') if err: return 1 self.file_elist = fname # have all options if self.make_3dd_contr: if not self.file_3dd_cmd: print '** cannot use -make_3dd_contrasts without -save_3dd_cmd' return 1 # advanced style has labels built in elif not self.labels and not self.advanced: print '** cannot use -make_3dd_contrasts without -stim_labels' return 1 if self.verb > 1: print '-- pre_stim_rest=%g, post_stim_rest=%g, seed=%s' \ % (self.pre_stim_rest, self.post_stim_rest, self.seed) print ' min_rest=%g, max_rest=%g,' \ % (self.min_rest, self.max_rest), \ 'offset=%g, t_gran=%g, t_digits=%d' \ % (self.offset, self.t_gran, self.t_digits) if self.labels: print ' labels are: %s' % ', '.join(self.labels) def has_any_opts(self, olist): """check user_opts for any name in olist""" for oname in olist: if self.user_opts.find_opt(oname): return 1 return 0 def apply_opt_timing_class(self, opt): """apply all -add_timing_class options usage: -add_timing_class label mindur [[mean max] [dtype [tgran]]] required: label mindur more: label mindur meandur maxdur all: label mindur meandur maxdur dtype tgran sample usage: -add_timing_class stimA 3 -add_timing_class stimA 3 5 10 -add_timing_class stimA 3 5 10 dist=decay -add_timing_class stimA 3 3 3 dist=decay t_grid=0.1 """ params = opt.parlist nparm = len(params) oname = '-add_timing_class' error_string = "** bad usage in '%s %s'" % (oname, ' '.join(params)) if nparm < 2: print error_string return 1 # get label (verify that it is not a number) name = params[0] try: ff = float(name) print error_string print ' first parameter should be a timing class label' return 1 except: pass # ------------------------------------------------------------ # get all durations try: min_dur = float(params[1]) except: print error_string print ' second parameter should be a duration' return 1 # just one duration implies all 3 if nparm == 2: mean_dur = max_dur = min_dur # two is not allowed elif nparm == 3: print error_string print " either supply a fixed duration or 'min mean max'" return 1 # get all 3 elif nparm >= 4: try: mean_dur = float(params[2]) max_dur = float(params[3]) except: print error_string print ' not all 3 durations convert to float' return 1 # ------------------------------------------------------------ # check non-decreasing, for any durs not equal to -1 # (min must be set) # and check for non-decreasing or negative (allow zero) if min_dur < 0: print error_string print " invalid negative duration" return 1 # meandur is allowed to be -1, for unspecified if mean_dur == -1: if self.verb > 2: print "-- timing class %s will have no fixed mean duration" % name elif min_dur > mean_dur: print error_string print " durations must have: min <= mean" return 1 # max can also be -1, for unspecified if max_dur == -1: if self.verb > 2: print "-- timing class %s will have no maximum duration" % name # max set, mean unset, so check min vs max elif mean_dur == -1 and min_dur > max_dur: print error_string print " durations must have: min <= max" return 1 # max set, mean set, so check mean vs max elif mean_dur != -1 and mean_dur > max_dur: print error_string print " durations must have: mean <= max" return 1 # ------------------------------------------------------------ # ready to roll, create the actual timing class instance # # just pass all other parameters, plus verb tclass = LRT.TimingClass(name, min_dur, mean_dur, max_dur, params=params[4:], verb=self.verb) if tclass.status: return 1 if self.verb > 1: print "++ adding new Timing class '%s'" % name if self.verb > 2: tclass.show('new timing class') self.tclasses.append(tclass) return 0 def apply_opt_stim_class(self, opt): """apply all -add_stim_class options usage: -add_stim_class label Nreps stim_class rest_class All parameters are required. """ params = opt.parlist nparm = len(params) oname = '-add_stim_class' error_string = "** bad usage in '%s %s'" % (oname, ' '.join(params)) if nparm != 4: print error_string return 1 # ------------------------------------------------------------ # check the 4 parameters: name, nreps, stname, rtname # get label (verify that it is not a number) name = params[0] try: ff = float(name) print error_string print ' first parameter should be a stim class label' return 1 except: pass # get nreps (verify that it is a positive integer) try: nreps = int(params[1]) except: print error_string print ' second parameter should be nreps (integer > 0)' return 1 # get stim and rest timing classes, verify existence stname = params[2] rtname = params[3] sclass = self.get_timing_class(stname) if sclass == None: print error_string print "** did not find timing class '%s' for stim" % stname return 1 rclass = self.get_timing_class(rtname) if rclass == None: print error_string print "** did not find timing class '%s' for rest" % rtname return 1 # ------------------------------------------------------------ # ready to roll, create the actual stim class instance sclass = LRT.StimClass(name, nreps, sclass, rclass, verb=self.verb) if sclass == None: print error_string return 1 if self.verb > 1: print "++ adding new Stim class '%s'" % name if self.verb > 2: sclass.show('new stim class') self.sclasses.append(sclass) return 0 def get_timing_class(self, name): """given name, return an LRT.TimingClass instance from tclasses if not found, return None """ for cc in self.tclasses: if cc.name == name: return cc return None def get_stim_class(self, name): """given name, return an LRT.StimClass instance from sclasses if not found, return None """ for cc in self.sclasses: if cc.name == name: return cc return None def labels_to_indices(self, labels, print_err=1): """convert the labels list into an index list - entries must be in self.labels return index_list, err (0 = success) """ if len(labels) == 0: return [], 0 if len(self.labels) == 0: if print_err: print '** missing labels for conversion' return [], 1 ilist = [] for lab in labels: try: ind = self.labels.index(lab) except: print '** assumed label %s is not in label list' % lab return [], 1 ilist.append(ind+1) return ilist, 0 def create_timing(self): """create stimulus timing files - for each run - compute random rest time as total run time minus stim time (sum of reps * time, per class) minus pre and post stimulus rest times - divide rest be gran to get number of rest trials - randomize: sequence of trials per class and rest - accumulate time: offset + gran (rest) or stim_dur - each class gets a list of times - write all timing lists to files """ if self.verb > 1: print '\n++ creating timing...' if self.seed != None: if self.verb > 1: print '++ init random with seed %d' % self.seed random.seed(self.seed) # possibly get timing across all runs at once if self.across_runs: self.stimes = self.make_rand_timing(self.num_runs, self.run_time[0]) if self.stimes == None: return 1 if len(self.stimes) != self.num_stim or \ len(self.stimes[0]) != self.num_runs: print '** make_rand_timing failure: bad list size' return 1 # create a timing list for each run separately else: # init the 3D array: class by run by stim # each element (class) is a list (run) of lists (stim) # (for each run, append a stim_dur list to each class) self.stimes = [[] for i in range(self.num_stim)] for run in range(self.num_runs): # get timing for current run stim_list = self.make_rand_timing(1, self.run_time[run]) # check for failure if stim_list == None: print '** rand_timing failure for run %d' % (run+1) return 1 if len(stim_list) != self.num_stim: print '** bad stim_list length %d' % len(stim_list) return 1 if len(stim_list[0]) != 1: print '** bad stim_list[0] length %d' % len(stim_list[0]) return 1 # add lists to stimes for index in range(self.num_stim): self.stimes[index].append(stim_list[index][0]) if self.verb > 2: print '--------- final list ---------' for s in range(len(self.stimes)): for r in range(len(self.stimes[0])): print UTIL.gen_float_list_string(self.stimes[s][r], '++ stim list[%d][%d] = '%((s+1),r)) # print self.stimes return None def set_filenames(self): """create a list of filenames for writing""" if self.prefix: prefix = self.prefix # be sure we have a prefix else: prefix = 'stim_times' self.fnames = [] for sind in range(self.num_stim): if self.labels: fname = '%s_%02d_%s.1D' % (prefix, sind+1, self.labels[sind]) else: fname = '%s_%02d.1D' % (prefix, sind+1) self.fnames.append(fname) def adv_write_timing_files(self): """write files using AfniData class""" if self.prefix: prefix = self.prefix # be sure we have a prefix else: prefix = 'stim_times' for sind, sc in enumerate(self.sclasses): sc.adata.fname = '%s_%02d_%s.1D' % (prefix, sind+1, sc.name) sc.adata.dur_len = -1.0 if sc.adata.write_as_timing(): return 1 return 0 def write_event_list(self, fname): """write event list to given file ('' or '-' or 'stdout' means stdout) self.full_event.list = per run list of [cind sdur tdur time] eall = [ [ [1 3.2] [4 0.7] .. ] [ [0 4.2] [2 3.1] .. ] [ [4 2.0] [2 1.9] .. ] ] """ if fname: if self.verb > 3: print "++ writing event list to file '%s'" % fname fd = open(fname, 'w') if not fd: print "** failed to open '%s' for writing event list" % fname return 1 else: fd = sys.stderr # note how much space will we need for '(NAME)' maxlen = 2 + max([len(sc.name) for sc in self.sclasses]) emptystr = '%*s' % (maxlen, '') if fd == sys.stderr: fd.write('# ==== all events ====\n') for rind, erun in enumerate(self.full_event_list): fd.write('-- run %d events --\n' % (rind+1)) fd.write('class %s start dur rest\n' % emptystr) fd.write('----- %s ----- ---- ----\n' % emptystr) for event in erun: nstr = '(%s)' % self.sclasses[event[0]].name fd.write('%3d %-*s %6.2f %6.2f %6.2f\n' \ % (event[0], maxlen, nstr, event[3], event[1], event[2])) fd.write('\n') if fd != sys.stderr: fd.close() return 0 def write_timing_files(self): """write timing from slist to files from the prefix""" if len(self.stimes) != self.num_stim or self.num_stim <= 0: print '** bad stim data for file write' return 1 if len(self.fnames) != self.num_stim: print '** missing filenames for timing output' return 1 # rcr - # if using AfniData class, then when writing timing: # if durations are constant, do not write them to file # to do this, set mtype to 0 before writing and reset # compute min and max stim times, for verbose output mint = max(self.run_time) maxt = 0 for sind in range(self.num_stim): stim_all = self.stimes[sind] # all times for this stim stim = sind+1 # 1-based index # open file, write each row (run), and close fname = self.fnames[sind] fp = open(fname, 'w') if not fp: print "** failed to open timing file '%s'" % fname return 1 if self.verb > 0: print 'writing timing file: %s' % fname for rind in range(self.num_runs): stims = stim_all[rind] if len(stims) == 0: # empty run, use '*' if rind == 0: fp.write('* *\n') else: fp.write('*\n') continue # write out all times on one line for ttime in stims: # if t_digits is not specified, use general format if self.t_digits < 0: fp.write('%.g ' % ttime) else: fp.write('%.*f ' % (self.t_digits, ttime)) if ttime < mint: mint = ttime if ttime > maxt: maxt = ttime if rind == 0 and len(stims) == 1: fp.write('*') fp.write('\n') fp.close() # and add this filename to the list, in case we want it later if self.verb > 1: print 'min, max stim times are: %.1f, %.1f' % (mint, maxt) return None def make_3dd_cmd(self): """write sample usage of 3dDeconvolve -nodata to a file""" if not self.file_3dd_cmd: return if os.path.isfile(self.file_3dd_cmd): print "** 3dD command file '%s' already exists, failing..." \ % self.file_3dd_cmd return if len(self.fnames) != self.num_stim: print '** fname list does not have expected length' return # set tr and nt for the command if self.tr != 0.0: tr = self.tr elif self.t_gran > 1.0: tr = self.t_gran else: tr = 1.0 nt = round(UTIL.loc_sum(self.run_time) / tr) cmd = '# -------------------------------------------------------\n' \ '# create 3dDeconvolve -nodata command\n\n' polort = UTIL.run_time_to_polort(self.run_time[0]) # separate the 3dDecon command, to apply wrappers c2 = '3dDeconvolve \\\n' \ + ' -nodata %d %.3f \\\n' % (nt, tr) \ + ' -polort %d \\\n' % polort \ + '%s' % make_concat_from_times(self.run_time,tr) \ + ' -num_stimts %d \\\n' % self.num_stim for ind in range(len(self.fnames)): c2 += ' -stim_times %d %s %s \\\n' % \ (ind+1,self.fnames[ind],basis_from_time(self.stim_dur[ind])) # add labels, but play it safe if self.labels and len(self.labels) == len(self.fnames): c2 += ' -stim_label %d %s \\\n' % (ind+1,self.labels[ind]) if self.make_3dd_contr and self.labels: c2 += self.make_3dd_contr_str(self.labels, prefix=' ') c2 += ' -x1D X.xmat.1D\n\n' first = (polort+1) * len(self.run_time) if len(self.fnames) > 1: # note first non-poly baseline index c2 += '# compute the sum of non-baseline regressors\n' \ "3dTstat -sum -prefix sum_ideal.1D X.xmat.1D'[%d..$]'\n\n" \ % first ynames = '-ynames SUM - sum_ideal.1D ' c2 += "# consider plotting the SUM below non-polort regressors\n"\ "# command: 1dplot -xlabel Time %sX.xmat.1D'[%d..$]'\n" \ % (ynames, first) else: c2 += "# consider plotting the desired regressor\n"\ "# command: 1dplot -xlabel Time X.xmat.1D'[%d]'\n" % first cmd += UTIL.add_line_wrappers(c2) if self.verb > 0: print "saving 3dD command to file '%s'...\n" % self.file_3dd_cmd if self.verb > 1: print cmd fp = open(self.file_3dd_cmd,"w") if not fp: print '** failed to open %s for writing 3dDecon cmd' \ % self.file_3dd_cmd, return fp.write(cmd) fp.close() def make_3dd_contr_str(self, labels, prefix=''): """return a string with all pairwise contrasts""" if not labels: return '' cstr = '' cind = 0 llist = labels # just to make shorter for first in range(len(llist)-1): for next in range(first+1,len(llist)): cstr += "%s-gltsym 'SYM: %s -%s' -glt_label %d %s-%s \\\n" % \ (prefix, llist[first], llist[next], cind+1, llist[first], llist[next]) cind += 1 if cind > 0: nstr = "%s-num_glt %d \\\n" % (prefix, cind) return nstr + cstr return '' def disp_time_stats(self): """display statistics of ISIs (inter-stimulus intervals)""" if not self.isi: if self.verb: print '-- no ISI stats to show' return print '\nISI statistics :\n\n' \ 'data min mean max stdev\n' \ '----------- ------- ------- ------- -------' print 'pre-rest %7.3f %7.3f %7.3f %7.3f' % \ (UTIL.min_mean_max_stdev(self.prerest)) print 'post-rest %7.3f %7.3f %7.3f %7.3f\n' % \ (UTIL.min_mean_max_stdev(self.postrest)) for ind in range(len(self.isi)): m0, m1, m2, s = UTIL.min_mean_max_stdev(self.isi[ind]) print 'run #%d ISI %7.3f %7.3f %7.3f %7.3f' % \ (ind, m0, m1, m2, s) allruns = [] for run in self.isi: allruns.extend(run) print '\nall runs ISI %7.3f %7.3f %7.3f %7.3f\n' % \ (UTIL.min_mean_max_stdev(allruns)) if self.verb > 3: for ind in range(len(self.isi)): print UTIL.gen_float_list_string(self.isi[ind],'ISI #%d : '%ind) print if self.verb > 3: for ind in range(len(self.isi)): self.isi[ind].sort() print UTIL.gen_float_list_string(self.isi[ind], 'sorted ISI #%d : '%ind) def make_rand_timing(self, nruns, run_time): """Create random timing for stimuli over all runs (or for a single run), where the random placement of stimuli is over all runs at once (so the number per run will probably vary). This can be called per run, or for all runs at once. nruns : number of runs run_time : total time, per run (in seconds) nstim : number of stimulus classes reps_list : number of repetitions per class (total over runs) sdur_list : duration of each stimulus class min_rest : minimum duration between any 2 stimuli offset : time offset to add to each stimulus time tinitial : initial rest time, per run tfinal : final rest time, per run ctrl_breaks : multi-runs: add max stim time to tfinal tgran : granularity of time, to partition rest verb : verbose level """ nstim = self.num_stim # number of stimulus types reps_list = self.num_reps # num reps of each stim class sdur_list = self.stim_dur # duration per class tinitial = self.pre_stim_rest tfinal = self.post_stim_rest min_rest = self.min_rest offset = self.offset ctrl_breaks = self.control_breaks tgran = self.t_gran verb = self.verb if verb > 2: print '-- make_rand_timing: nruns = %d' % nruns print ' rtime, offset, tinit, tfinal = %g, %g, %g, %g' \ % (run_time, offset, tinitial, tfinal) print ' reps_list = %s' % reps_list print UTIL.gen_float_list_string(sdur_list,' sdur_list = ') # print ' sdur_list = %s' % sdur_list print ' tgran = %.4f, verb = %d' % (tgran, verb) # verify inputs if nruns <= 0: print '** make_rand_timing error: nruns = %d' % nruns return if run_time <= 0.0 or nstim <= 0 or tinitial < 0 or tfinal < 0: print '** bad rand_timing inputs: rtime, nstim, tinit, tfinal = '+ \ '%g, %d, %g, %g' % (run_time, nstim, tinitial, tfinal) return if tgran < gDEF_MIN_T_GRAN: print '** time granularity (%f) below minimum (%f)' % \ (tgran, gDEF_MIN_T_GRAN) if not reps_list or not sdur_list or \ len(reps_list) != nstim or len(sdur_list) != nstim: print '** invalid rand_timing input lists: reps, stimes = %s %s '% \ (reps_list, sdur_list) return # steal a copy of sdur_list so that we can trash it with min_rest stim_durs = sdur_list[:] # if multi runs and ctrl_breaks, add max stim time (-tgran) to tfinal if nruns > 1 and ctrl_breaks: smax = max(sdur_list) if verb > 1: print '++ adding max stim (%.1f) to post_stim_rest' % smax tfinal += smax # compute total stim time across all runs together # min_rest is essentially added to each stimulus time stime = 0.0 for ind in range(nstim): if reps_list[ind] < 0 or stim_durs[ind] < 0: print '** invalid negative reps or stimes list entry in %s, %s'\ % (reps_list, stim_durs) return stime += reps_list[ind]*stim_durs[ind] # compute rest time across all runs together tot_rest = nruns * run_time - stime isi_rtime = nruns * (run_time - tinitial - tfinal) - stime # account for min_rest, and apply with stim_durs if min_rest < 0.0: min_rest = 0.0 rtime = isi_rtime - min_rest * UTIL.loc_sum(reps_list) nrest = int(rtime / tgran) if verb > 1 or self.show_timing_stats: print '\n++ total time = %.1f, (stim = %.1f, rest = %.1f)\n' \ ' init rest = %.1f, end rest = %.1f, min rest = %.1f\n' \ ' total ISI = %.1f, rand ISI = %.1f\n' \ ' rand ISI rest time = %d intervals of %.3f seconds\n' % \ (nruns*run_time, stime, tot_rest, tinitial, tfinal, min_rest, isi_rtime, rtime, nrest, tgran) if rtime == 0: print '** warning, exactly no time remaining for rest...' elif rtime < 0: print '** required stim and rest time exceed run length, failing...' return # make a random events list, based on reps_list[] and nrest err, elist = self.randomize_events(nrest) if err: return if self.max_rest > 0: if self.apply_max_rest(elist): return # convert the event list into a list of stimulus lists slist = [[[] for i in range(nruns)] for j in range(nstim)] # keep some statistics s_isi = [] # isi's, per run s_first = 1 # first event, per run s_laststim = 0 # end of last stimulus (in current run) ctime = tinitial run = 0 for event in elist: # maybe we are done with this run if ctime >= run_time - tfinal: # statistics s_first = 1 # have first event in this run self.isi.append(s_isi) # store previous list s_isi = [] self.postrest.append(run_time - s_laststim) s_laststim = 0 # normal work run += 1 ctime = tinitial if run >= nruns: # we've gone beyond our bounds print '** failure! total run time has been exeeded...' return if not event: ctime += tgran # rest event: tgran seconds else: # statistics if s_first: # then first event in this run self.prerest.append(ctime) s_first = 0 else: s_isi.append(ctime - s_laststim) eind = event - 1 # event is one more than index atime = ctime+offset # add any requested offset slist[eind][run].append(atime) # append cur time to event list ctime += stim_durs[eind] # increment time by this stim s_laststim = ctime # store the stim end time ctime += min_rest # and rest # and set end-of-run stats self.postrest.append(run_time - s_laststim) self.isi.append(s_isi) if verb > 3: print '-- end make_random_timing() --' for s in range(nstim): for r in range(nruns): print UTIL.gen_float_list_string(slist[s][r], '++ stim list[%d][%d] = '%((s+1),r)) return slist def valid_orderstim(self): """- every index in {1..num_stim} can occur only once over all orderstim lists - only values {1..num_stim} can occur in orderstim lists - for each olist, make sure num_reps is constant return 1 if valid, 0 if not""" if len(self.orderstim) < 1: return 1 used = [] errs = 0 for olist in self.orderstim: if self.verb > 4: print "-- testing olist: %s" % olist if len(olist) < 2: print "** orderstim too short: %s" % olist errs += 1 for val in olist: if val < 1 or val > self.num_stim: print "** orderstim list has values outside {1..%d}: %s" \ % (self.num_stim, olist) errs += 1 elif val in used: print "** orderstim '%d' is not unique in lists" % val errs += 1 else: used.append(val) if errs: continue # and check that all reps are the same vlist = [self.num_reps[v-1] for v in olist] if not UTIL.vals_are_constant(vlist, cval=None): print "** stimuli in order list need constant nreps\n" \ " orderstim: %s\n" \ " nreps: %s\n" % (olist, vlist) errs += 1 if errs: if self.verb > 1: print '** valid_orderstim: %d errors' % errs return 0 else: if self.verb > 1: print '-- %d orderstim lists are valid' % len(self.orderstim) return 1 def randomize_ordered_events(self, nrest): """create a list of randomized events, subject to orderstim - return error code (0=success) and event list """ if len(self.orderstim) < 1: print '** no orderstim to apply' return 1, None if not self.valid_orderstim(): return 1, None # add anything that is not used used = [] for olist in self.orderstim: used.extend(olist) elist = [] nstim = self.num_stim # ------------------------- step 1 --------------------------- # catenate lists of event types that are either not in orderstim # or are the first element in one # fill elist with unused stimuli for stim in range(nstim): sval = stim+1 # stim val, for ease of typing if sval not in used: if self.verb > 1: print '++ filling unordered stim %d' % sval elist += [sval for i in range(self.num_reps[stim])] if self.verb > 3: print '++ elist (len %d, unordered): %s' % (len(elist), elist) # and append 'first' entries to elist for ind in range(len(self.orderstim)): olist = self.orderstim[ind] sval = olist[0] elist += [sval for i in range(self.num_reps[sval-1])] if self.verb > 1: print '++ filling order[0] stim %d' % sval if self.verb > 3: print '++ elist (len %d, order base): %s' % (len(elist), elist) # ------------------------- step 2 --------------------------- # randomize this list (other orderlist stimuli are not random) UTIL.shuffle(elist) if self.verb > 3: print '++ elist (len %d, base, rand): %s' % (len(elist), elist) # ------------------------- step 3 --------------------------- # fill in orderstim events: # insert orderstim list at each orderstim initial element # given: elist, firstdict, orderstim ==> create new elist # create 'firstdict' lookup table for follower insertion firstdict = self.make_ordered_firstdict(self.orderstim) elist = self.insert_order_followers(elist, firstdict, self.orderstim) # ------------------------- step 4 --------------------------- # randomly merge with rest: # - shuffle binary list of rest/stim # - for each stim val, replace with respective stim index stat, mergelist = self.merge_elist_with_rest(elist, nrest) if stat: return 1, None return 0, mergelist def merge_elist_with_rest(self, elist, nrest): """make a new list with rest elements randomly dispersed along elist """ mergelist = [0 for i in range(nrest)] + [1 for i in range(len(elist))] UTIL.shuffle(mergelist) eind = 0 # counter for filling from elist for mind in range(len(mergelist)): if mergelist[mind]: mergelist[mind] = elist[eind] eind += 1 if eind != len(elist): print '** randomize panic: mergelist/elist mismatch' return 1, None return 0, mergelist def make_ordered_firstdict(self, orderstim): """return a dictionary of orderstim indices, indexed by first stims""" firstdict = {} for ind in range(len(orderstim)): olist = self.orderstim[ind] sval = olist[0] firstdict[sval] = ind # which list contains sval if self.verb > 3: print '++ firstdict lookup table: %s' % firstdict return firstdict def insert_order_followers(self, elist, firstdict, orderstim): """fill in followers from orderstim list inputs: elist : current list of first and unordered elements firstdict: dict of {first order element: index into orderstim} orderstim: list of ordered stim lists output: new elist: with followers inserted """ enew = [] for val in elist: if firstdict.has_key(val): # then append orderlist enew.extend(orderstim[firstdict[val]]) else: enew.append(val) if self.verb > 3: print '++ elist with followers (len %d, all stim): %s' \ % (len(enew), enew) return enew def randomize_events(self, nrest): """make a list of all event types and then rest, randomize the list - this may be subject to lists of ordered stimuli (orderstim) (stimuli must be unique across all such lists) - return error code (0=success) and event list""" if nrest < 0: print "** randomize_events: nrest < 0" return 1, [] # might also handle orderstim case if len(self.max_consec) == self.num_stim: return self.randomize_limited_events(nrest) if len(self.orderstim) > 0: return self.randomize_ordered_events(nrest) nstim = self.num_stim reps_list = self.num_reps elist = [] for stim in range(nstim): elist += [(stim+1) for i in range(reps_list[stim])] elist.extend([0 for i in range(nrest)]) if self.verb > 2: print '++ elist (len %d, sorted): %s' % (len(elist), elist) UTIL.shuffle(elist) if self.verb > 2: print '++ elist (len %d, random): %s' % (len(elist), elist) for stim in range(nstim): print ' number of type %d = %d' % (stim+1,elist.count(stim+1)) return 0, elist def randomize_limited_events(self, nrest): """return a random list of events, where each event type may have a limit on consecutive occurances - keep list of # remaining for each type (include rest) - fraction of # remaining provides probability for next event - do not include event type at limit - this should succeed until one event type remains which has reached its limit - then try to dispurse remaining events - finally, merge in rest """ # if stimuli are ordered, clear num_reps for followers if len(self.orderstim) > 0: if self.verb > 1: print '-- have limited events w/ordered stim...' num_reps = self.num_reps[:] for orderlist in self.orderstim: # stim in lists are 1-based, num_reps_is 0-based for follower in orderlist[1:]: num_reps[follower-1] = 0 else: num_reps = self.num_reps rv, clist = self.make_limited_space_list(num_reps, self.max_consec) # next, rewrite as elist ---> convert to 1-based index list here elist = [] for entry in clist: elist.extend([entry[0]+1 for i in range(entry[1])]) if self.verb > 2: print "== clist (0-based) %s" % clist if self.verb > 2: print "== elist (1-based) %s" % elist # if orderstim, create a firstdict table and insert follower events if len(self.orderstim) > 0: if self.verb > 1: print '-- inserting order followers...' firstdict = self.make_ordered_firstdict(self.orderstim) elist = self.insert_order_followers(elist, firstdict, self.orderstim) # finally, merge with rest elist = UTIL.random_merge(elist, [0 for i in range(nrest)]) return 0, elist def make_limited_space_list(self, num_reps, max_consec): """given a list of num_reps per stim_class and max_consec for each, return an event count list of the form [[eind, count] ...] return status, clist """ # initial pass on creation of events list rv, clist, rtype, rcount = \ self.init_limited_event_list(num_reps, max_consec) if rv: print '** failure of randomize_limited_events' return 1, None # if we ran out of space for one event type, try to fill # prev must be remaining event type if rcount > 0: rv, clist = self.fill_remaining_limited_space(max_consec, clist, rtype, rcount) if rv: return 1, None if self.verb > 4: print '++ MLSL: reps list %s' % clist return 0, clist def init_limited_event_list(self, num_reps, max_consec): """first pass on limited event list: Given num_reps and max_consec of each type, create a list [[event, count], ...] of event and consecutive counts. Make this sequentially, only appending what is possible given the limits. The probability of each new event will be the relative fraction of remaining events of that type, or 0 if the limit has been reached. inputs: num_reps and max_consec outputs: status - success or no from function clist - event count list [[event, count]] rtype - remaining type rcount - remaining count Note clist has 0-based event indices, they should be incremented. """ if self.verb > 3: print '-- init_limited_event_list\n' \ ' num_reps = %s\n' \ ' max_consec = %s' % (num_reps, max_consec) remain = num_reps[:] nremain = sum(remain) t_remain = remain[:] # same, but might get 0 at some index # convert max_consec, change any 0 to num_reps max_consec = max_consec for i, m in enumerate(max_consec): if m <= 0: max_consec[i] = num_reps[i] if self.verb > 1: print "-- limited events reps = %s, max = %s" % (remain, max_consec) # create a list of [[event, count]], from which to generate elist # (so insertion of over-the-limit events will be easier) # # note t_remain = remain, except that if the previous event type # hit its consec limit, temporarily set that t_remain to 0 clist = [[-1, 0]] # init with no rest events, say cind = 0 prev = -1 for ind in range(nremain): # if we are at a limit, do not let limited event happen if prev >= 0: if clist[cind][1] >= max_consec[prev]: t_remain[prev] = 0 t_nremain = sum(t_remain) if t_nremain == 0: # we are either completely done, or should move on to insertion # of remaining 'prev' type events break # so insert a new event # make eind the next chosen event type; it must be usable rind = int(t_nremain*random.random()) # in [0,t_nremain-1] eind = 0 while eind < len(t_remain) and rind >= t_remain[eind]: rind -= t_remain[eind] eind += 1 # verify that we didin't screw up if eind >= len(t_remain): print '** eind too big for t_remain, failing...' return 1, None, 0, 0 # account for applied event index and resync t_remain remain[eind] -= 1 t_remain[eind] -= 1 if prev >= 0: t_remain[prev] = remain[prev] # modify clist and monitor limit if eind == prev: # if same as previous, increment clist[cind][1] += 1 else: # else, have a new event type clist.append([eind, 1]) cind += 1 prev = eind nremain -= 1 if remain[prev] != nremain: print '** LE failure, nremain = %d, remain = %s' % (nremain,remain) return 1, None, 0, 0 if not self.limited_events_are_valid(clist, max_consec): print '** LE failure, some events exceed maximum' return 1, None, 0, 0 if self.verb > 1: if nremain == 0: print '-- have 0 events to insert' else: print '++ have %d events of type %d to insert' % (nremain, prev) print ' (note: events are currently 0-based)' return 0, clist, prev, nremain def fill_remaining_limited_space(self, max_consec, clist, rtype, rcount): """Try to add rcount events of type rtype into clist, subject to limits imposed by max_consec. Note that there should not be space for rtype at the end of the list (else it would have already been added). Note that if inserting an rtype event between two events of other types, then there is actually space for max_consec[rtype] events in that position. """ pmax = max_consec[rtype] if clist[-1] != [rtype, pmax]: print "** messed up max for limited events..." return 1, None # do we have enough space for remaining events? space, npos = self.count_limited_space(clist, rtype, max_consec[rtype]) if space < rcount: # we are in trouble print '** limited_events: only %d positions for %d inserts' \ % (space, rcount) return 1, None if self.verb > 2: print '-- space for insert: %d, positions %d ' % (space, npos) if self.verb > 3: print ' clist %s' % clist # okay, insert the remaining rcount events of type rtype # rely on rcount being small and do linear searches for each insert for ind in range(rcount): # rcount times, insert rtype event rind = int(npos*random.random()) # insertion index if self.verb > 2: print '-- random insertion index %d of %d' % (rind, npos) ecount = 0 # count of all potential events # find valid position index rind (check positions per cind) for cind in range(1, len(clist)): c0, c1 = clist[cind][0], clist[cind][1] # if rind is too big to insert here, move on # (take care if rtype bin is full) # if rtype, either no postions or c1+1 postions available # (consider after last as one more position) if c0 == rtype: if c1 >= pmax: continue # no postions if rind >= c1+1: rind -= c1+1 continue # else, this is good # if prev is rtype, one fewer position elif clist[cind-1][0] == rtype: if rind >= c1-1: rind -= (c1-1) continue # else, this is good; add 1 to skip first element rind += 1 # else can go before any elif rind >= c1: rind -= c1 continue # we have a winner! break and insert break if cind >= len(clist): print '** LE: failed to find insertion index' return 1, None if self.verb>3: print '++ inserting at index %d (of %d), rind %d' \ % (cind, len(clist), rind) # if adjacent to an 'rtype' event, increment event count if c0 == rtype: clist[cind][1] += 1 # if at beginning, insert a new event elif rind == 0: clist.insert(cind, [rtype, 1]) else: # must break clist[cind] apart # decrement by rind, insert new event, insert prev rind clist[cind][1] -= rind # decrement and insert by rind clist.insert(cind, [rtype, 1]) clist.insert(cind, [c0, rind]) # fill cases subtraced pmax+1, so all can add 1 here space -= 1 if self.verb > 5: print ' clist %s' % clist snew, npos = self.count_limited_space(clist, rtype, max_consec[rtype]) if space != snew: print "** space count failure, space = %d, count = %d" \ % (space, snew) return 1, None if not self.limited_events_are_valid(clist, max_consec): print '** LE fill remain failure, some events exceed maximum' return 1, None return 0, clist def count_limited_space(self, clist, eind, emax): """Given an event class list, and event index and the maximum number of sequential events of each type, return the number of such events that could possibly be added, as well as the number of positions available to insert a new event. Note: eind must be last event type. """ space = 0 positions = 0 if self.verb > 5: print '== eind, emax: %d, %d' % (eind, emax) print '== CLS clist: %s' % clist for ind in range(1, len(clist)): if self.verb > 5: print ' %02d %s ' % (ind, clist[ind]), # if next is eind, can go up to max ncur = clist[ind][1] # note number of current entries if clist[ind][0] == eind: space += emax - ncur if emax - ncur > 0: positions += ncur if self.verb > 5: print space, positions continue # have ncur postions times emax space, unless prev was eind space += ncur*emax positions += ncur # now adjust if prev was eind: # no additional space before first event # if prev is full, position cannot be before first event if clist[ind-1][0] == eind: space -= emax if clist[ind-1][1] == emax: positions -= 1 if self.verb > 5: print space, positions # maybe there is space at the end ind = len(clist)-1 if clist[ind][0] != eind: print '** space at end (index %d of %d), this should not happen' \ % (ind, len(clist)) print '== clist: %s' % clist space += emax positions += 1 return space, positions def limited_events_are_valid(self, clist, maxlist): """verify that each event type in clist does not have too many in a row return 1 if valid, 0 if not""" for ind in range(1, len(clist)): # if next is full, skip if clist[ind][1] > maxlist[clist[ind][0]]: print '** check events failure for maxlist %s' % maxlist print '%s' % clist return 0 return 1 def apply_max_rest(self, eventlist): """modify eventlist so that there is never more than max_rest rest before any event""" if self.max_rest <= 0: return max_rest = self.max_rest if self.min_rest > 0: max_rest -= self.min_rest if self.verb > 2: print '-- updating max_rest from %g to %g' \ % (self.max_rest, max_rest) # note the maximum rest as a number of intervals maxnrest = int(max_rest/self.t_gran) # create a list of stim index, post-stim rest count, stim type lists rlist = [] elist = [-1, 0, 0] # pre-event rest 'event' rlist.append(elist) for ind in range(len(eventlist)): val = eventlist[ind] if val > 0: # non-rest event, add new elist to rlist elist = [ind, 0, val] rlist.append(elist) else: elist[1] += 1 if self.verb > 3: print '++ event rlist', rlist # make extendable and fix lists elist = [] # extend list (space to add rest) flist = [] # fix list (need to remove space) tot = 0 for ind in range(len(rlist)): if rlist[ind][1] > maxnrest: flist.append(ind) tot += (rlist[ind][1] - maxnrest) elif rlist[ind][1] < maxnrest: elist.append(ind) if self.verb > 3: print '-- maxnrest = %d (%s/%s = %s)' \ % (maxnrest, max_rest, self.t_gran, max_rest/self.t_gran) print '-- maxrest extendable lists:', elist print '-- maxrest fix lists:', flist fl = [rlist[flist[i]][1]-maxnrest for i in range(len(flist))] el = [maxnrest-rlist[elist[i]][1] for i in range(len(elist))] print '===========================================================' print 'fix : %s' % ' '.join(['%s' % v for v in fl]) print '===========================================================' print 'ext : %s' % ' '.join(['%s' % v for v in el]) print '===========================================================' # maybe there is nothing to do if len(flist) < 1: if self.verb > 1: print '-- no event block exceeds max rest' return # note how much rest we need to move around ftotal = UTIL.loc_sum([rlist[flist[i]][1]-maxnrest for i in range(len(flist))]) etotal = UTIL.loc_sum([maxnrest-rlist[elist[i]][1] for i in range(len(elist))]) if self.verb > 1: print '-- shifting %g seconds of excessive rest (%g available)' \ % (ftotal*self.t_gran, etotal*self.t_gran) if self.verb > 3: print '== excessive rest exeeds %d events...' % maxnrest print ' (%s/%s = %s)' % (max_rest, self.t_gran, max_rest/self.t_gran) if etotal < ftotal: print "** max_rest too small to adjust (%g < %g)" % (etotal, ftotal) return elif self.verb > 3: print '--> distributing %d rest events among %d slots' \ % (ftotal, etotal) for fix in flist: nshift = rlist[fix][1] - maxnrest for count in range(nshift): if len(elist) <= 0: print '** panic: empty elist but rest to shift' sys.exit(1) # choose where to shift rest to and shift it rind = int(len(elist)*random.random()) # shift: insert (at rlist[elist[rind]]+1 and remove roff = rlist[elist[rind]][0]+1 # new rest offset eventlist[roff:roff] = [0] # insert the rest for iind in range(elist[rind]+1, len(rlist)): # adjust rlist rlist[iind][0] += 1 # maybe we are out of rest for this elist element rlist[elist[rind]][1] += 1 if rlist[elist[rind]][1] >= maxnrest: elist[rind:rind+1] = [] # then remove from elist # delete the extra rest from this event rlist[fix][1] -= nshift roff = rlist[fix][0] + 1 # rest to delete eventlist[roff:roff+nshift] = [] # actual deletion # and drop fix rest down by nshift for iind in range(fix+1, len(rlist)): # adjust rlist rlist[iind][0] -= nshift if self.verb > 3: print '++ fixed event rlist', rlist if self.verb > 4: print '++ updated eventlist', eventlist etotal = UTIL.loc_sum([maxnrest-rlist[elist[i]][1] for i in range(len(elist))]) print '-- updated etotal = %d' % etotal # ====================================================================== # ADV: functions for modern timing method def adv_create_timing(self): """create stimulus timing event list create list of events (array of event type/duration) - put 2D list of married data elements in StimClass (mdata element = [time [AM list] duration], [t [] d]) fill with rest do for each run separately rest for last stim might be zero """ if self.verb > 1: print '\n++ new: creating timing...' if self.seed != None: if self.verb > 1: print '++ init with random seed %d' % self.seed random.seed(self.seed) # create stimdict, a name -> StimClass index list for sind, sc in enumerate(self.sclasses): self.stimdict[sc.name] = sind if self.verb > 3: print '++ have len %d stimdict : %s' \ % (len(self.stimdict.keys()), self.stimdict) # apply any such command line options to the stim classes if self.set_max_consec(): return 1 # durations are created per stim class even if the timing types are # the same (so more time in one class does not cost time in another) if LRT.create_duration_lists(self.sclasses, self.num_runs, self.across_runs, verb=self.verb): return 1 # simlpy randomize order of all events per run if self.adv_randomize_event_lists(): return 1 # add rest, return final event list if self.adv_create_afnidata_list(): return 1 if self.file_elist or self.verb > 4: if self.write_event_list(self.file_elist): return 1 return 0 def set_max_consec(self): """apply max_consec list here return 0 on success, 1 on error""" nmc = len(self.max_consec) nsc = len(self.sclasses) if nmc == 0 or nsc == 0: return 0 if nmc == 1: # expand list to linclude all self.max_consec = [self.max_consec[0]] * nsc elif nmc != nsc: print '** have %d -max_consec params, but %d stim classes'%(nmc,nsc) return 1 mclist = self.max_consec mcnames = [self.sclasses[ind].name for ind in range(nmc) \ if mclist[ind] > 0] print '++ applying max_consec to: %s' % ', '.join(mcnames) if len(mcnames) == 0: return 0 # apply to each stim class for sind, sc in enumerate(self.sclasses): if mclist[sind] <= 0: continue if sc.max_consec > 0: print '** max_consec alread set for class %s' % sc.name return 1 sc.max_consec = mclist[sind] if self.verb > 1: print '-- setting max_consec = %d for class %s' \ % (sc.max_consec, sc.name) return 0 def adv_randomize_event_lists(self): """For each run (or across), put all events in a list and randomize. return an array of run lists of [stimind duration], e.g. eall = [ [ [1 3.2] [4 0.7] .. ] [ [0 4.2] [2 3.1] .. ] [ [4 2.0] [2 1.9] .. ] ] """ if self.across_runs: ntodo = 1 else: ntodo = self.num_runs ordered = self.valid_ordered_stim() if ordered < 0: return 1 eall = [] for rind in range(ntodo): # either applying max_consec or not if len(self.max_consec) == len(self.sclasses): erun = self.adv_limited_shuffled_run(rind, ordered) if erun == None: return 1 else: erun = [] for sind, sc in enumerate(self.sclasses): # if ordered stim and sc is not a leader: skip if ordered and sc.name in self.osfollow: if self.verb>3: print '-- no shuffle for follower %s' % sc.name continue erun.extend([[sind, dur] for dur in sc.durlist[rind]]) UTIL.shuffle(erun) if self.verb > 1: print '-- randomized event lists (no followers) for run %d' % rind self.disp_consec_event_counts([erun]) # if ordered stim, insert followers if ordered: erun = self.adv_insert_followers(rind, erun) eall.append(erun) self.stim_event_list = eall if self.across_runs: if self.adv_partition_sevents_across_runs(): return 1 if self.verb > 4: print '-- randomized event lists across all runs' self.disp_consec_event_counts(eall) return 0 def disp_consec_event_counts(self, eall): # first convert to a list of counts per class, across runs numc = len(self.sclasses) call = [[] for sind in range(numc)] for erun in eall: eind = 0 elen = len(erun) while eind < elen: ecur = eind eind += 1 cind = erun[ecur][0] while eind < elen: if erun[eind][0] != cind: break eind += 1 # have (eind-ecur) events of type cind call[cind].append(eind-ecur) if self.verb > 1: print '++ consec event counts:' slen = max([len(sc.name) for sc in self.sclasses]) if self.verb > 4: for sind, sc in enumerate(self.sclasses): print '-- consec list for #%2d=%-*s (len %3d): %s' \ % (sind, slen, sc.name, len(call[sind]), call[sind]) print for sind, sc in enumerate(self.sclasses): mi, mn, mx, st = UTIL.min_mean_max_stdev(call[sind]) if self.verb > 1: print ' consec for %*s, sum %4d,' \ ' mmms: %7.3f %7.3f %7.3f %7.3f' \ % (slen, sc.name, sum(call[sind]), mi, mn, mx, st) def adv_limited_shuffled_run(self, rind, ordered): """return randomized events subject to max_consec - if ordered stimuli, do not include followers """ numc = len(self.sclasses) if len(self.max_consec) != numc: print '** ALSR: bad max_consec list: %s' % self.max_consec return None # create a modified reps list that omits followers reps = [len(sc.durlist[rind]) for sc in self.sclasses] if ordered: for sind, sc in enumerate(self.sclasses): if sc.name in self.osfollow: if self.verb>3: print '-- no consec for follower %s' % sc.name reps[sind] = 0 # generate a random list of the form [[sind nconsec] ...] rv, clist = self.make_limited_space_list(reps, self.max_consec) if rv: return None clist.pop(0) # remove initial [-1,0] # convert to form [[sind dur] ...] enew = [] for cind, consec in enumerate(clist): # append consec[1] durations of type consec[0] sind = consec[0] sc = self.sclasses[sind] for tind in range(consec[1]): dur = sc.durlist[rind].pop(0) enew.append([sind, dur]) return enew def adv_insert_followers(self, rind, events): """given list of [sind, dur], insert appropriate followers return new event list""" if self.verb > 3: print '-- insert followers: starting with %d events' % len(events) enew = [] for event in events: ename = self.sclasses[event[0]].name enew.append(event) # possibly insert followers if ename in self.osleaders: for fname in self.osdict[ename]: sind = self.stimdict[fname] sc = self.sclasses[sind] dur = sc.durlist[rind].pop(0) enew.append([sind, dur]) if self.verb > 3: print '-- insert followers: returning %d events' % len(enew) return enew def valid_ordered_stim(self): oname = '-ordered_stimuli' nostim = len(self.orderstim) nolabs = len(self.orderlabs) if nostim > 0 and nolabs == 0: print '** advanced usage requires labels in %s' % oname return -1 if nolabs == 0: return 0 errs = 0 alllabs = [] leaders = [] followers = [] for olabs in self.orderlabs: if len(olabs) < 2: print '** %s requires at least 2 labels, have: %s' \ % (oname, ' '.join(olabs)) errs += 1 continue lead = olabs[0] follow = olabs[1:] for lab in olabs: if lab not in self.labels: print '** invalid label %s in %s %s' \ % (lab, oname, ' '.join(olabs)) return -1 # fatal if lab in alllabs: print '** %s: label %s used multiple times' % (oname, lab) errs += 1 alllabs.append(lab) # number of events must match ls = self.sclasses[self.stimdict[lead]] for fname in follow: fs = self.sclasses[self.stimdict[fname]] if fs.nreps != ls.nreps: print '** -ordered_stimuli: classes must have equal num stim' print ' %s has %d, %s has %d' \ % (lead, ls.nreps, fname, fs.nreps) errs += 1 self.osdict[lead] = follow leaders.append(lead) followers.extend(follow) if errs: return -1 self.osleaders = leaders self.osfollow = followers if self.verb > 2: print '-- have %d ordered stim, with %d leaders and %d followers' \ % (nolabs, len(leaders), len(followers)) if self.verb > 3: print ' leaders : %s' % ', '.join(leaders) print ' followers: %s' % ', '.join(followers) return 1 def adv_partition_sevents_across_runs(self): """break stim_event_list into num_runs""" if not self.across_runs: return 0 print '** rcr - adv_partition_sevents_across_runs' return 1 # in across_runs case, we do not yet know run breaks def adv_create_afnidata_list(self): """create AfniData instances per stim class (decide on rest, which implies all timing) (across_runs: nothing special to do anymore) convert stim_events to full_events (i.e. attach times, base on rest) create corresponding mdata lists per stim class create corresponding AfniData instances """ # add rest to convert stim_event_list to full_event_list if self.adv_create_full_event_list(): return 1 # convert to AfniData instances if self.adv_create_adata_list(): return 1 return 0 def adv_create_adata_list(self): """convert full_event_list to AfniData objects given a list of [sind sdur rdur stime] per run make a list of mdata = [stime [] sdur] per run, per sclass convert mdata to AfniData object, stored as adata in stim class - """ # add any offset offset = self.offset if offset != 0 and self.verb > 1: print '++ applying offset of %g to all times' % offset for sind, sc in enumerate(self.sclasses): sc.mdata = [] for rind, erun in enumerate(self.full_event_list): srun = [event for event in erun if event[0] == sind] mrun = [[se[3]+offset, [], se[1]] for se in srun] sc.mdata.append(mrun) sc.adata = LAD.AfniData(mdata=sc.mdata, verb=self.verb) sc.adata.name = sc.name return 0 def adv_create_full_event_list(self): """given stim_event_list of form: [ [ [sind dur] [sind dur] ... ] [ [sind dur] [sind dur] ... ] ] create full_event_list (append rest dur and event time) of form: [ [ [sind sdur rdur stime] [sind sdur rdur stime] ... ] [ [sind sdur rdur stime] [sind sdur rdur stime] ... ] ] return 0 on success """ # input selist, output felist selist = self.stim_event_list felist = [] for rind, erun in enumerate(selist): rv, ferun = self.adv_stim2full_run_elist(rind, erun) if rv: return 1 if self.verb > 5: print '== full timing list: run %d, elist:' % (rind+1) for fe in ferun: print ' %s' % ', '.join(['%g'%e for e in fe]) felist.append(ferun) self.full_event_list = felist return 0 def adv_stim2full_run_elist(self, rind, events): """convert an event_list line: [ [sind dur] [sind dur] ... ] ] into a full_event_list line (includes rest dur and stim time): [ [sind sdur rdir time] [sind sdur rdir time] ... ] - add pre-stim and post-stim rest events, in case they have needed timing classes - compute total rest time (run_time - sum of stim durs) rest time includes pre/post stim rest - make list of all used rest classes (for this run) - partition total rest time across rest classes - distribute across applied rest classes return status, full_event_list """ # quick checks: nevents = len(events) if nevents == 0: return 0, events elif nevents == 1: events[0].append(self.pre_stim_rest) return 0, events # total stim time before pre-/post-stim rest events are included stime = sum([e[1] for e in events]) # ----------------------------------------------------------------- # include pre- and post- stimulus rest events : # dur entry comes from options # rest entry comes from timing class # (these rest events come and go just within this function) # also, self.rand_post_stim_rest affects rest for last stim # rcr - explain in help events.insert(0, [-2, self.pre_stim_rest]) events.append([-1, self.post_stim_rest]) pprest = self.pre_stim_rest+self.post_stim_rest rtime = self.run_time[rind] - stime randtime = rtime-pprest if self.verb > 2 or randtime < 0: print '-- run %02d: run time = %s, stime = %s, rtime = %s' \ % (rind+1, self.run_time[rind], stime, rtime) print ' pre-rest = %s, post-rest = %s, random rest = %s' \ % (self.pre_stim_rest, self.post_stim_rest, randtime) if randtime < 0: print '** unsolvable random timing for rest' return 1, [] rv, rcounts, rtypes = self.count_all_rest_types(events) if rv: return 1, events rv, rtimes = self.partition_rest_time(randtime, rcounts, rtypes) if rv: return 1, events if self.verb > 3: print '== partitioned rest:' for cind, rc in enumerate(rtypes): print " %4d rest events of type '%s', time = %g" \ %(rcounts[cind], rc.name, rtimes[cind]) # get rest events, and apply to timing (append accumulated time) for rtind, rc in enumerate(rtypes): rc.etimes = LRT.random_duration_list(rcounts[rtind], rc, rtimes[rtind], force_total=1) # add option, show rest details if self.verb > 5 or self.show_rest_events: mesg='run %d rest' % rind rc.show_durlist_stats(rc.etimes, mesg=mesg, details=1) # quick test if self.verb > 3: rc = sum([len(rc.etimes) for rc in rtypes]) ec = len(events) print '== have %d rest events and %d stim events' % (rc, ec) # for each event, get the current rest time rall = 0 try: for eind, event in enumerate(events): sind = event[0] if eind == nevents and not self.rand_post_stim_rest: # no post-stim rest (this is not in any rest list, # since it does not match rest class for stim) # (also, #nevents is before inserting pre-stim rest) rc = g_instant_timing_class elif sind == -2: rc = self.pre_stimc elif sind == -1: rc = self.post_stimc else: rc = self.sclasses[event[0]].rclass rtime = rc.etimes.pop(0) event.append(rtime) except: print '** rest time extraction error' return 0, events # replace rest time with event time ctime = 0 for event in events: # append current time, and increment by stim and rest times event.append(ctime) ctime += event[1] + event[2] # and finally remove the pre- and post- rest events if events[0][0] != -2 or events[-1][0] != -1: print '** pre/post-stim rest events out of order' else: events.pop(0) events.pop() return 0, events def partition_rest_time(self, rtot, rcounts, rtypes): """given counts and corresponding rest types, partition total rest time per rest type - try to allocate min time for all types - for all types with means, decide on total needed time (above min) - if there is not quite enough time, rescale their times - for now, extra rest will trickle to end return status, rtimes """ ntypes = len(rtypes) # ----------------------------------------------------------------- # compute and distribute min_time tot_min = 0 rtimes = [0] * ntypes for rind, rc in enumerate(rtypes): # compute min total, store it and accumuate mt = rc.min_dur * rcounts[rind] rtimes[rind] = mt tot_min += mt remain = rtot - tot_min if remain < 0: print '** partition rest: insufficient space for even min rest' return 1, [] # now go back and track those with applied means (mean > min) # - also, note whether we can alter the times later tot_mean = 0 have_rand = 0 # have unrestricted rest can_alter = [0] * ntypes # we can adjust times for these classes for rind, rc in enumerate(rtypes): offset = rc.mean_dur - rc.min_dur if offset != 0: can_alter[rind] = 1 # do we have classes with unrestricted rest? if offset < 0: have_rand = 1 # now skip anything without known space if offset <= 0: continue # total requested remaining time is mean * nevents mt = offset * rcounts[rind] tot_mean += mt # ----------------------------------------------------------------- # check for early exit if remain == 0 (warn user if not complete) # (remain < 0 was checked above, but be safe) if remain <= 0: if tot_mean > 0: print '** PRT: no time left for above-min mean-based rest' if have_rand == 1: print '** PRT: no time left for above-min random rest' return 0, rtimes # ----------------------------------------------------------------- # possibly scale down offsets for means mean_scalar = 1 if tot_mean > remain: mean_scalar = remain * 1.0 / tot_mean # or, scale UP if there are no max-less classes elif have_rand == 0 and tot_mean > 0: mean_scalar = remain * 1.0 / tot_mean # possibly provide more details on scalar if abs(mean_scalar-1) > 0.05 or self.verb > 2: print '-- have %g of %g mean seconds available, scaling by %g' \ % (remain, tot_mean, mean_scalar) if remain > tot_mean and have_rand: print ' (no scaling due to unlimited rest classes)' # ----------------------------------------------------------------- # now actually distribute into rtimes # min time is done, so next distribute postive mean time if tot_mean > 0: for rind, rc in enumerate(rtypes): offset = rc.mean_dur - rc.min_dur if offset <= 0: continue mt = mean_scalar * offset * rcounts[rind] rtimes[rind] += mt remain -= tot_mean # for have_rand, partition remaining time based on number of events if remain > 0 and have_rand > 0: nevents = 0 # first get total number of events for rind, rc in enumerate(rtypes): if rc.mean_dur < rc.min_dur: nevents += rcounts[rind] # then partition time based on fractional number of events for rind, rc in enumerate(rtypes): if rc.mean_dur < rc.min_dur: rtimes[rind] = remain * rcounts[rind] * 1.0 / nevents # any undistributed rest will trickle to post-stim rest return 0, rtimes def count_all_rest_types(self, elist): """given an event list of [etype, dur] elements, get a unique list of rest timing classes and their counts return status, [rcount], [rtype] """ rcounts = [] rtypes = [] nume = len(elist) for eind, event in enumerate(elist): # there are special cases for which rest class to use if eind == 0: # pre-stim rest event rc = self.pre_stimc elif eind == (nume-1): # post-stim rest event rc = self.post_stimc elif eind == (nume-2) and not self.rand_post_stim_rest: # use INSTANT rest after last event rc = g_instant_timing_class else: # ahhh, the usual case, get rest class out of stim class sc = self.sclasses[event[0]] rc = sc.rclass # if we already have this one, increment the count, else append if rc in rtypes: rind = rtypes.index(rc) rcounts[rind] += 1 else: rtypes.append(rc) rcounts.append(1) if self.verb > 2: print "-- have %d rest events across %d rest classes" \ % (sum(rcounts), len(rcounts)) if self.verb > 3: for cind, rc in enumerate(rtypes): print " %4d rest events of type '%s'"%(rcounts[cind],rc.name) return 0, rcounts, rtypes def adv_rest_across_runs(self, sall): print '** RCR - rest across runs...' return [] def adv_make_3dd_cmd(self): """write sample usage of 3dDeconvolve -nodata to a file""" if not self.file_3dd_cmd: return if os.path.isfile(self.file_3dd_cmd): print "** 3dD command file '%s' already exists, failing..." \ % self.file_3dd_cmd return nstim = len(self.sclasses) # set tr and nt for the command if self.tr != 0.0: tr = self.tr elif self.t_gran > 1.0: tr = self.t_gran else: tr = 1.0 nt = round(UTIL.loc_sum(self.run_time) / tr) cmd = '# -------------------------------------------------------\n' \ '# create 3dDeconvolve -nodata command\n\n' polort = UTIL.run_time_to_polort(self.run_time[0]) # separate the 3dDecon command, to apply wrappers c2 = '3dDeconvolve \\\n' \ + ' -nodata %d %.3f \\\n' % (nt, tr) \ + ' -polort %d \\\n' % polort \ + '%s' % make_concat_from_times(self.run_time,tr) \ + ' -num_stimts %d \\\n' % self.num_stim for ind, sc in enumerate(self.sclasses): basis = adv_basis_from_time(sc) # if duration modulation as stim durs vary, use AM1 if basis == 'dmUBLOCK': c2 += ' -stim_times_AM1 %d %s %s \\\n' % \ (ind+1, sc.adata.fname, basis) else: c2 += ' -stim_times %d %s %s \\\n' % \ (ind+1, sc.adata.fname, basis) c2 += ' -stim_label %d %s \\\n' % (ind+1,sc.name) # and possibly add contrasts if self.make_3dd_contr: labels = [sc.name for sc in self.sclasses] c2 += self.make_3dd_contr_str(labels, prefix=' ') if self.prefix: xmat = 'X.%s.xmat.1D' % self.prefix else: xmat = 'X.xmat.1D' c2 += ' -x1D %s\n\n' % xmat cmd += UTIL.add_line_wrappers(c2) # add additional commands for plotting c2 = '' first = (polort+1) * len(self.run_time) tlabel = '-xlabel Time' if len(self.sclasses) > 1: # note first non-poly baseline index c2 += '# compute the sum of non-baseline regressors\n' \ "3dTstat -sum -prefix sum_ideal.1D %s'[%d..$]'\n\n" \ % (xmat, first) ynames = '-ynames SUM -' c2 += "# consider plotting the SUM below non-polort regressors:\n"\ '# (if(0) allows easy copy-and-paste)\n' \ "if ( 0 ) then\n" \ " 1dplot -sepscl %s %s sum_ideal.1D %s'[%d..$]'\n" \ "endif\n\n" \ % (tlabel, ynames, xmat, first) # include a more detailed version, just for kicks tlabel = '-xlabel "Time index"' ynames = '-ynames SUM %s -' \ % ' '.join([sc.name for sc in self.sclasses]) c2 += "# or the more completely labeled command:\n" \ "if ( 0 ) then\n" \ " 1dplot -sepscl %s \\\n" \ " %s \\\n " \ " sum_ideal.1D %s'[%d..$]'\n" \ "endif\n\n" \ % (tlabel, ynames, xmat, first) else: c2 += "# consider plotting the desired regressor\n"\ "# command: 1dplot %s %s'[%d]'\n" % (tlabel, xmat, first) c2 += '\n' cmd += UTIL.add_line_wrappers(c2) if self.verb > 0: print "saving 3dD command to file '%s'...\n" % self.file_3dd_cmd if self.verb > 1: print cmd if UTIL.write_text_to_file(self.file_3dd_cmd, cmd): print '** failed to write 3dD command to %s' % self.file_3dd_cmd, return def adv_basis_from_time(sclass): const, cdur = sclass.adata.check_constant_duration() if const: if cdur > 1: return "'BLOCK(%g,1)'" % cdur return 'GAM' else: return 'dmUBLOCK' def basis_from_time(stim_len): if stim_len > 1.0: return "'BLOCK(%g,1)'" % stim_len else: return 'GAM' def make_concat_from_times(run_times, tr): if len(run_times) < 2: return '' str = " -concat '1D: 0" cind = 0 for ind in range(len(run_times)-1): cind += round(run_times[ind]/tr) str += ' %d' % cind return str + "' \\\n" def do_isi_pdfs(argv): oname = '-show_isi_f_pdf' if oname in argv: ind = argv.index(oname) try: NT = int(argv[ind+1]) NR = int(argv[ind+2]) except: print '** %s requires 2 integer params, NTask, NRest' % oname return 1 show_sum_pswr(NT, NR) oname = '-show_isi_pdf' if oname in argv: ind = argv.index(oname) try: NT = int(argv[ind+1]) NR = int(argv[ind+2]) except: print '** %s requires 2 integer params, NTask, NRest' % oname return 1 show_isi_pdf(NT, NR) return 0 def show_isi_pdf(T, R): """akin to show_sum_pswr, but init and accumulate: P(R=0) = T/(T+R) P(R=r+1)/P(R=r) = (R-r)/(R+T-1-r) """ pcur = 1.0*T/(T+R) cump = pcur rat = pcur print 'nstart prob inc' print '------ ---------- ----------' for r in range(0,R+1): print "%5d %-10g %-10g" % (r, pcur, rat) rat = (1.0*R - r) / (R + T - 1 - r) pcur *= rat cump += pcur print 'cum result is %g' % cump def show_sum_pswr(nT, nR): cp = 0.0 prev = 0 print 'nstart prob inc' print '------ ---------- ----------' for r in range(nR+1): p = prob_start_with_R(nT,nR,r) cp += p # print 'prob at %3d = %g (cum %g)' % (r, p, cp) if prev == 0: prev = p print r, p, p/prev prev = p print 'cum result is %g' % cp def prob_start_with_R(nA, nB, nS): """return the probability of starting nS (out of nB) class B elements should equal: choose(nB, nS)*nS! * nA *(nB+nA-nS-1)! / (nA+nB)! or: factorial(nB, init=nB-nS+1) * nA / fact(nA+nB, init=nA+nB-nS) or: choose(nB,nS)/choose(nA+nB,nS) * nA/(nA+nB-nS) """ return 1.0 * nA * UTIL.factorial(nB, init=nB-nS+1) \ / UTIL.factorial(nA+nB, init=nA+nB-nS) def process(): timing = RandTiming('make random timing') timing.init_opts() rv = timing.read_opts() if rv != None: # 0 is okay, else error code if rv: UTIL.show_args_as_command(sys.argv,"** failed command:") return rv rv = timing.process_opts() if rv != None: if rv: UTIL.show_args_as_command(sys.argv,"** failed command:") return rv # new, advanced way if len(timing.sclasses) > 0: rv = timing.adv_create_timing() if rv: UTIL.show_args_as_command(sys.argv,"** failed command:") return 1 if timing.adv_write_timing_files(): return 1 if timing.file_3dd_cmd: timing.adv_make_3dd_cmd() # ignore return value if timing.show_timing_stats: LAD.show_multi_isi_stats(timing.sclasses, timing.run_time, timing.tr, verb=(timing.verb>2)) # old way else: rv = timing.create_timing() if rv != None: if rv: UTIL.show_args_as_command(sys.argv,"** failed command:") return rv rv = timing.set_filenames() if rv != None: if rv: UTIL.show_args_as_command(sys.argv,"** failed command:") return rv rv = timing.write_timing_files() if rv != None: if rv: UTIL.show_args_as_command(sys.argv,"** failed command:") return rv if timing.file_3dd_cmd: timing.make_3dd_cmd() # ignore return value if timing.show_timing_stats: timing.disp_time_stats() # ignore rval return 0 if __name__ == "__main__": rv = process() sys.exit(rv)