-
Notifications
You must be signed in to change notification settings - Fork 177
Description
Describe the bug
I am very new to using SimBA, so I'm sorry if this is frequently asked.
I'm trying to import a .h5 analysis file that I generated through SLEAP, but I am getting this error: OSError: Unable to synchronously open file (bad object header version number)
When I try it with a .slp file (the original project folder), it starts going through the frames (yay!), but it gets stuck once it finished and throws this:
Restructuring SLEAP frame: 353284/353285, Video: 20251208_TS1_Gr3_T1
SIMBA DATA HEADER ERROR: SIMBA ERROR: SimBA expects 48 columns of data inside the files within project_folder/csv/input_csv directory. However, within file 20251208_TS1_Gr3_T1 file, SimBA found 420 columns. Thus, there is 372 more data columns in the imported data than anticipated, which may represent 124 bodyparts if each body-part has an x, y and p value. Either revise the SimBA project pose-configuration with 124 more body-part, or include 124 less body-part in the imported data
So a similar issue I imagine?
I've been trying to open a .csv output, however, my videos are very large with a lot of labeled parts (353000 frames, 10 parts), so the file is 2.6 GB and takes a long time to open (I had it running for an hour and it still hadn't opened :/) . I have a lot of trials to go through, so this isn't very efficient.
To Reproduce
Steps to reproduce the behavior:
- Create a project config as normal
- Import a CSV (SLEAP) file in the project settings window
Expected behavior
I expect the program to open with my predictions from sleap and I can start classifying behaviors.
Screenshots
Unable to open .h5 file:
Unable to open .slp file
Desktop (please complete the following information):
Windows 11
Python 3.8
SimBA installed in a Anaconda environment in version 4.5.5
I am using SLEAP 1.4.1
Additional context
Add any other context about the problem here.