Data Prep

Before run_everything is useful:

1. get smr files

• rename them accordingly. Should be mousename_side_trial#.smr
  • ie gp14m1_left_1.smr
  • note that sometimes the side is missing, this is ok.
  • recommended to use Thunar bulk renaming utility on Linux
• should be in data_batch/experimental_group/smr directory
  • ie Full_data/WT_Levelwalk/smr directory

2. batch convert smr to h5. This will generate a folder called h5_raw. It will also generate phase files
3. move phase files to their own directory by hand. That directory should be named phase.
4. Sanity Check: Look at the phase files + fix the knee joints in locoproc. Save in dir called h5_knee_fixed (easiest is just to rename h5_raw to h5_knee_fixed; OTHERWISE re-running batch-convert erases your knee joint corrections).

• advised to use separate virtual environment for locoproc b/c of numpy versions
• swing is red, stance is green. Otherwise, should fix in excel/sheets/libreoffice calc
  • recommend LibreOffice spreadsheet application, Excel is bad for this
  • MUST be 'swing' followed by 'stance' column, should delete first swing stance if needed so that times align
    • This set up is required for locoproc to read it
  • MUST be swing time followed by stance time also
    • NOTE: This will feel 'backwards', but it is REQUIRED for locomotion-analysis

It is likely worth writing something to test that the phase file format is fixed & that auto-corrects deleting the first stance time if needed.