video_frame_ind not upsampled in TrodesPosV1

[samuelbray32]

Describe the bug

• TrodesPosV1 gives option to upsample position data
• However, video_frame_ind is not upsampled in this process
• Causes two errors:
  1. video_frame_ind is not accurate map to position (only an issue if making a video)
  2. Creates mismatch between number of timesteps and data size in the nwb object nwb.scratch['BehavioralTimeSeries']['video_frame_ind']. This violates Dandi standards and prevents validation/upload during export (see error below, ratio of timestamps to data size is same as upsampled to original rate).

Severity.ERROR: The length of the first dimension of data (36303) does not match the length of timestamps (605029). in /stelmo/nwb/export/sept2024_testing/dandiset_sept2024_testing/Yoshi20220518_2AJRJC0ATL.nwb

Solutions New Files

• Fix the bug in the TrodesV1 code by upsampling the frame index

Existing Files

• I would prefer not to delete and re-populate these entries, as this would require re-doing a good amount downstream.
• Can write a script to edit the nwb and resolve database checksum