Open
Description
Requirement Text
Explore using Tensorflow Timeseries Generator feature for LSTM __pre_process_data
Background Information
LSTM currently preprocesses data manually. https://www.tensorflow.org/api_docs/python/tf/keras/preprocessing/sequence/TimeseriesGenerator provides tools for doing this
see also keras.preprocessing.timeseries_dataset_from_array() from https://keras.io/examples/timeseries/timeseries_weather_forecasting/
Note- those above are deprecated- look for alternative
Something like this: https://www.tensorflow.org/api_docs/python/tf/keras/utils/timeseries_dataset_from_array
Suggested Solution
https://www.tensorflow.org/api_docs/python/tf/keras/utils/timeseries_dataset_from_array
DoD
- Explore if TimeseriesGenerator will work and is appropriate
- Implement TimeseriesGenerator, if determined appropriate