Dataset file and Excel date conversion in json_lat/json_lon columns

[maren-ha]

Hi, thanks for publishing the R-for-eaRth challenge materials.

I think there might be a problem with the dataset file currently included in the repository. The notebook useR_preprocessing.ipynb loads:

file_path = 'useR_sentinel2_indices_mean.csv'
df = pd.read_csv(file_path, sep=';')

but the repository currently appears to provide only useR_sentinel2_indices_mean.xlsx.

Some values in the json_lat and json_lon columns of the xlsx appear to have been converted to Excel dates. Some rows include values shown as:

29. Jul
Jun 91
31. Aug
06. Nov
Jul 53
05. Sep
Oct 64

I've tried to reconstruct the decimal values behind them and I think I got them right because I could eventually reproduce the figure shown in the example notebook, but just to make sure:
Would it be possible to upload the original useR_sentinel2_indices_mean.csv used in the notebook?

Thanks a lot!

[pbiecek]

Thank you for pointing this out
I've added the original data in csv file to https://github.com/CredibleAI/R-for-eaRth/blob/main/useR_sentinel2_indices_mean.csv
now it should work

[maren-ha]

Great, thank you very much! Yes it works.