Remove time_30 and time_5

[JackKelly]

After #223 is implemented, the code should be able to handle any arbitrary sample rate. And, once we output each modality into a different NetCDF file, each NetCDF file will have a single sample period. So always refer to time as time and remove any code that talks about time_5 or time_30

[peterdudfield]

This is covered in
#229

[peterdudfield]

This is now merged, do shout if there is something you think is not done @JackKelly

[JackKelly]

Cool beans. Do we still need these methods in nowcasting_dataset.config.model.Process:

• seq_length_30_minutes()
• seq_length_5_minutes()
• history_minutes_divide_by_30()
• forecast_minutes_divide_by_30()

[peterdudfield]

Im doing them in #217

but i think we do, or at least some of them. the sequence thing is useful for making test data.

The validation on history minutes, could only apply to gsp data for example. Or maybe ive mis understood how it will be used

[JackKelly]

cool beans, sounds good!