This repository folder contains a few scripts for both SPONT and SPR.
You should start by creating a folder for each and copying the relevant scripts into that folder.
Navigate to either the SPONT or SPR folder. Then, run the optimize script:
sh optimize.sh
The optimize script randomly generates a bunch of timing files based on the length of your experiment, the TR length, the number of stimulus types, and the number of trials per stimulus type. It then calculates the efficiency of each iteration of these timing files.
This should give you a file called results.txt. The columns in results include iteration, eff, and seed. Iteration is the index for the iteration in that row. Eff refers to efficiency -- lower numbers indicate a more efficient (better) design. Seed is the randomly-generated number used to randomize and generate the stimulus timing files.
results.txt automatically has a line at the bottom indicating which iteration is the most efficient. However, you can also open results.txt in Excel and sort by eff to see which designs are the most efficient. This might be neccessary if the most efficient design doesn't have the most optimal stimulus order (e.g., lots of the same stimulus type in a row).
Once you've run optimize.sh and looked at results.txt, you can use testiter.sh to create timing files for a given iteration suitable for experimental presentation software.
To run, testiter.sh, type the command followed by the iteration number you'd like to generate the timing file for.
- E.g.,
sh testiter.sh
This wil create a two files (one for each run; SPONT_timings_run#.csv or SPR_timings_run#.csv) that has onset times and stimulus types for a given iteration. You should present a stimulus with the type indicated in TrialType at each Onset Time.