Add the trainer back to the forecaster after saving model to file

[ziqin8]

🔬 Background

• Why is this change needed? Is there a related issue or a new feature to be added?

After save, forecaster.predict won't work as the trainer is removed.

To reproduce:

import pandas as pd
from neuralprophet import NeuralProphet, set_log_level, df_utils
from neuralprophet import save as npsave

data_location = "https://raw.githubusercontent.com/ourownstory/neuralprophet-data/main/datasets/yosemite_temps.csv"
df = pd.read_csv(data_location)

# first model
m = NeuralProphet()
_ = m.fit(df, epochs=2, learning_rate=0.01, progress=None)
npsave(m, "test.np")
forecast = m.predict(df.iloc[-2:])

---------------------------------------------------------------------------
AttributeError                            Traceback (most recent call last)
Cell In[110], line 12
     10 _ = m.fit(df, epochs=2, learning_rate=0.01, progress=None)
     11 npsave(m, "test.np")
---> 12 forecast = m.predict(df.iloc[-2:])

File ~/mambaforge/envs/neuralprophet/lib/python3.8/site-packages/neuralprophet/forecaster.py:1049, in NeuralProphet.predict(self, df, decompose, raw)
   1047 forecast = pd.DataFrame()
   1048 for df_name, df_i in df.groupby("ID"):
-> 1049     dates, predicted, components = self._predict_raw(
   1050         df_i, df_name, include_components=decompose, prediction_frequency=self.prediction_frequency
   1051     )
   1052     df_i = df_utils.drop_missing_from_df(
   1053         df_i, self.config_missing.drop_missing, self.predict_steps, self.n_lags
   1054     )
   1055     if raw:

File ~/mambaforge/envs/neuralprophet/lib/python3.8/site-packages/neuralprophet/forecaster.py:2746, in NeuralProphet._predict_raw(self, df, df_name, include_components, prediction_frequency)
   2744     self.model.set_covar_weights(self.model.get_covar_weights())
   2745 # Compute the predictions and components (if requested)
-> 2746 result = self.trainer.predict(self.model, loader)
   2747 # Extract the prediction and components
   2748 predicted, component_vectors = zip(*result)

AttributeError: 'NoneType' object has no attribute 'predict'

🔮 Key changes

• Explain the main changes introduced by this pull request for the reviewer.

Add the trainer back to the forecaster after torch.save

📋 Review Checklist

• I have performed a self-review of my own code.
• I have commented my code, added docstrings and data types to function definitions.
• I have added pytests to check whether my feature / fix works.

Please make sure to follow our best practices in the Contributing guidelines.

[codecov]

Codecov Report

Merging #1416 (6a4a070) into main (1c6094d) will increase coverage by 0.00%.
The diff coverage is 100.00%.

@@           Coverage Diff           @@
##             main    #1416   +/-   ##
=======================================
  Coverage   89.87%   89.88%           
=======================================
  Files          38       38           
  Lines        5066     5071    +5     
=======================================
+ Hits         4553     4558    +5     
  Misses        513      513           

Files Changed	Coverage Δ
neuralprophet/utils.py	79.65% <100.00%> (+0.30%)	⬆️

📢 Have feedback on the report? Share it here.

[leoniewgnr]

This is really nice! Thanks @c3-ziqin for this change!

[github-actions]

Model Benchmark

Benchmark	Metric	main	current	diff
YosemiteTemps	MAE_val	1.34899	1.34899	0.0%	✅
YosemiteTemps	RMSE_val	2.00817	2.00817	0.0%	✅
YosemiteTemps	Loss_val	0.00078	0.00078	0.0%	✅
YosemiteTemps	MAE	1.32133	1.32133	0.0%	✅
YosemiteTemps	RMSE	2.13713	2.13713	0.0%	✅
YosemiteTemps	Loss	0.00064	0.00064	0.0%	✅
YosemiteTemps	time	57.5074	60.93	5.95%	⚠️
PeytonManning	MAE_val	0.58162	0.58162	0.0%	✅
PeytonManning	RMSE_val	0.72218	0.72218	0.0%	✅
PeytonManning	Loss_val	0.01239	0.01239	0.0%	✅
PeytonManning	MAE	0.41671	0.41671	0.0%	✅
PeytonManning	RMSE	0.55961	0.55961	0.0%	✅
PeytonManning	Loss	0.00612	0.00612	0.0%	✅
PeytonManning	time	12.3769	12.71	2.69%	✅
AirPassengers	MAE_val	13.0627	13.0627	0.0%	✅
AirPassengers	RMSE_val	15.9453	15.9453	0.0%	✅
AirPassengers	Loss_val	0.00131	0.00131	0.0%	✅
AirPassengers	MAE	9.88153	9.88153	0.0%	✅
AirPassengers	RMSE	11.7354	11.7354	0.0%	✅
AirPassengers	Loss	0.00052	0.00052	0.0%	✅
AirPassengers	time	5.16755	5.36	3.72%	⚠️

Model training plots

Model Training

PeytonManning

YosemiteTemps

AirPassengers