This PR changes the format of the BPF files that are stored by PhaseVocoder and FaceWarp to make them directly compatible with deployment in the JONES/revcor app (issue #18)
Why
The revcor online experimentation app expects:
that parameters are stored in filenames with the same basename as the stimuli and the .txt extension
file format with a header (which is used to get column names in the dataframe storing experiment's results) and
comma-separated.
PhaseVocoder's fixed behaviour so far has files as "_bpf.txt", no header, and tab separated. FaceWarp has header and comma-separated, and the ability to change file ext, but default is ".dfm"
How
added a create_BPF_header function in bpf.py that assembles header str depending on transform and config
modified BPF save to include header and comma-separated
added a param_ext option in config file syntax, which is now read by PhaseVocoder and replaces the old dfm_ext option in FaceWarp
changed default FaceWarp ext to .txt
set default PhaseVocoder ext to .txt
added documentation on param_ext
added param_ext to the tutorial config files
Testing
Tested using the speech and face tutorial, checking correct file format and names in single calls & chained
What
This PR changes the format of the BPF files that are stored by PhaseVocoder and FaceWarp to make them directly compatible with deployment in the JONES/revcor app (issue #18)
Why
The revcor online experimentation app expects:
PhaseVocoder's fixed behaviour so far has files as "_bpf.txt", no header, and tab separated.FaceWarphas header and comma-separated, and the ability to change file ext, but default is ".dfm"How
create_BPF_headerfunction inbpf.pythat assembles header str depending on transform and configPhaseVocoderand replaces the olddfm_extoption inFaceWarpFaceWarpext to .txtPhaseVocoderext to .txtTesting
Tested using the speech and face tutorial, checking correct file format and names in single calls & chained