Error in Train model

[sgoldenlab]

Error in Train model

[JensBlack]

This is a print out from the function simba/train_model_2.py just before the try (mentioned above).

print(data_train)
        Deviation_from_median_Movement_difference_movement_center_movement_left_hip  ...  movement_tail_tip
89150                                            0.548807                            ...           4.247683
104100                                          -0.112851                            ...           0.638911
56938                                            0.541995                            ...           1.495396
74278                                            0.249147                            ...           0.541398
69342                                            0.345994                            ...           0.886768
...                                                   ...                            ...                ...
48366                                           -1.046650                            ...           2.202447
27243                                            0.826896                            ...           1.022565
79840                                            0.674166                            ...           1.100300
35520                                            0.391709                            ...           2.451280
83918                                            0.488544                            ...           1.342929

[91256 rows x 763 columns] 

print(target_train)
[1. 1. 1. ... 1. 1. 1.]

print(len(target_train))
91256 

[JensBlack]

Update:
Removing the additional csv file (by deleting it in the target_inserted folder) and keeping the original csv file did not raise the error.
Removing the original csv file (but keeping the additional) raised the error.

Therefore I will add information about my procedure of adding the additional csv file:

1. Import csv file using the "import single csv file" in the "further imports" tab.
2. Copying video and frames (manually into the corresponding folders)
3. setting video parameters
4. outlier correction (parameter set to 1000; above resolution): No Correction
5. feature extraction
6. Addition of label columns for 3 models in both csv files via custom script (result see above)
7. Set Settings and train model

[JensBlack]

i solved the issue.... it was an error in my labelling method. I will close the issue.

For future reference:
The number of rows in the labelled columns did not match the number of frames in the csv file.
interestingly:
this is not represented in data_train and and target_train length, which both have the same length.

[benlansdell]

Hi @JensBlack, I may be having the same issue as you. What was the error you were making in your labeling method?

Thanks.

[benjaws]

Hi @benlansdell,
I have the same problem, did you solved it ?

Many thanks