ValueError: num_samples should be a positive integer value, but got num_samples=0

[LightingMc]

Hey,
So I downloaded the git and downloaded the data set into a subfolder inside the git. The installation of echonet went fine. But I get this error no matter I try to run any of the given functions.

I have tried renaming the downloaded folder with the .csv files and the videos, and have also edited that echonet.cfg file. I don't really understand what is going on.

[douyang]

This is an issue with file path, please change the .cfg to where you expect the videos. If videos are not present at the expected path, num_samples will return 0.

[LightingMc]

Hey, thanks for the reply and sorry for bothering you so much. I tried two approaches.

1. So in this git's repo I created a new folder with the name of a4c-video-dir. It has all of the videos.

I also copy-pasted the two .csv files into this folder

But I still get the same error.

2. Created another folder with the name of a4c-video-dir which had three items in it: the 2 csv files and a folder named Videos with all the videos in it. Still got the same errors

[LightingMc]

and I also tried these two above methods by editing the echonet.cfg file with a new folder name and tried methods 1,2 described before. Still got the same errors.

[douyang]

We've noticed a few errors like this when using a version of pytorch and torchvision different from requirements.txt. Please try using those versions and let us know if still getting that error.

[LightingMc]

Hey,
Thanks a lot for reaching back. I tried to install a conda env. from the requirements given. I did:

conda create --name Dynamic --file requirements.txt

But it doesn't work and gives me the following error:

Do you know how I could resolve this issue?

[LightingMc]

I am trying to do your thing. But need to know something. What is the file/tree structure of the a4c-video-dir?

For me doing ls a4c-video-dir gives FileList.csv Videos VolumeTracings.csv, where the Videos folder contains all of the videos.

[LightingMc]

The issue was resolved. The issue was that in the data provided to us via box, the .csv files that I got had .avi after the video file names. Your code for reading the videos assumes that they didn't have .avi after the video file names.

[douyang]

Thanks so much for the update! We will update the CSV to reflect that.

[atanuGuin]

@Beep-Boop-Bopa, could you please let me know your echonet.cfg configuration? All the videos and 2 csv files are kept at path:
/workspace/xxx/yyy/dynamic/a4c-video-dir
and my echonet.cfg containes the DATA_DIR path as follows:
(my_syft_env) root@380b6b959170:/workspace/xxx/yyy/dynamic# cat echonet.cfg
DATA_DIR =/workspace/xxx/yyy/dynamic/a4c-video-dir/

Thanks in advance.

[LightingMc]

In my experience, the echonet.cfg file thing doesn't really work that well. So I am not using it at all. I don't have a .config file. The python file which does "import echonet" is in the same directory as a4c-video-dir.

Right now, my code has .avi in it and is working. I dunno why.

[LightingMc]

Just try with and without the .avi extension in the csv files and see what works.

[chinmayabv]

Hi Im running into some issues and unable to figure out the contents of the csv file. Can you please provide me with a sample FileList.csv or the headers that need to be in it?

[chahalinder0007]

@chinmayabv
It seems you are not having the split column in your csv, the column determines if we are going to use a particular example for training validation or test
The sample csv looks like this:

Filename,EF,ESV,EDV,FrameHeight,FrameWidth,FPS,Frame,Split
0X100009310A3BD7FC.avi,78.49840597,14.88136797,69.21053366,112,112,50,174,VAL
0X1002E8FBACD08477.avi,59.10198811,40.38387624,98.74288352,112,112,50,215,TRAIN
0X1005D03EED19C65B.avi,62.36379841,14.2677839,37.90973395,112,112,50,104,TRAIN
0X10075961BC11C88E.avi,54.54509675,33.14308352,72.91420979,112,112,55,122,TRAIN
0X10094BA0A028EAC3.avi,24.88774165,127.5819446,169.8550242,112,112,52,207,VAL
0X100CF05D141FF143.avi,55.95179045,47.45383119,107.7315779,112,112,50,248,TEST
0X100E3B8D3280BEC5.avi,36.57373876,81.74728111,128.8855429,112,112,38,101,TRAIN
0X100E491B3CD58DE2.avi,33.99174499,76.86556916,116.4484187,112,112,50,226,TRAIN
0X100F044876B98F90.avi,69.31525203,25.13140943,81.90195812,112,112,50,221,TRAIN