input dataframe column is obliged to be named "ds", "y" and output dataframe column is olbliged to be named "yhat1"

[changhyeonnam]

As far as I know, to use neuralprophet, you need to write the following preprocessing code to rename the column name.

total_data.rename(columns={'timestamp': 'ds'}, inplace=True)
total_data.rename(columns={'value': 'y'}, inplace=True)

I know it's a convention of neuralprophet ('ds', 'y'), but it could be beneficial if you could add a feature that could change the column name like this.

self.model.fit(train_data, ds='timestamp', y='value', freq="5min")

so that, you don't need to add renaming method for column name.