Inference with a custom dataset

[MatthewBeddows]

Inference with a custom dataset

I downloaded the required data using "get_data_pl_short_length.py" and "get_data_sl_short_length.py".
For the sake of testing I have tried with the default time intervals and dates in these files.

I have then used "parallel_copy_small_set.py" to process these into my hdf5 file.

I can't see anything else I should modify in "inference.py" with this setup.

I graphed the predictions at a location in Scotland
latitude = 56.573492
longitude = -2.550597

Working Predictions
Using the provided dataset for inference I get good results:

Custom Inference Predictions
Here are the predictions using a custom dataset at inference (Using the default timeframes in the provided files) :

Anyone have any idea where I am going wrong with this?