When I run the global model, there's an issue with the ID column in the returned results—it's not displaying correctly

[stanhaishan]

When I run the global model, there's an issue with the ID column in the returned results—it's not displaying correctly.
code:
m = NeuralProphet(
trend_global_local="local",
season_global_local="local",
changepoints_range=0.8,
trend_reg=5,
n_forecasts=5,
n_lags=20,
unknown_data_normalization=True,
)
m.set_plotting_backend("plotly")

df_train, df_test = m.split_df(df1, valid_p=0.2, local_split=True)
metrics = m.fit(df_train, freq='D')
future = m.make_future_dataframe(df2,n_historic_predictions=True)
forecast = m.predict(future,decompose=False,raw=True)

[iago-figueiredo-a3]

I got the same problem. It seems like it's a bug with the raw=True parameter. If you set raw=False, the ID column seems right. Then you have to shift back the yhat to get the step behavior of n-days-ahead prediction