[minor] Memory efficient float32 types instead of float64 types

[leoniewgnr]

🔬 Background

The whole computation was done with float64 types, but torch tensors are float32 anyway and most of the ML models use float32 values. This will affect model accuracy but I think its worth it because we safe a lot of memory!

🔮 Key changes

• y and yhat values: float32 if float32 is inserted, float64 if float64 is inserted
• time values: will always be float64 because np.datetime64
• feature values (events): float32

[github-actions]

Model Benchmark

Benchmark	Metric	main	current	diff
YosemiteTemps	MAE_val	1.3442	1.34899	0.36%	✅
YosemiteTemps	RMSE_val	2.00245	2.00817	0.29%	✅
YosemiteTemps	Loss_val	0.00077	0.00078	0.58%	✅
YosemiteTemps	MAE	1.3192	1.32133	0.16%	✅
YosemiteTemps	RMSE	2.13518	2.13713	0.09%	✅
YosemiteTemps	Loss	0.00064	0.00064	0.17%	✅
YosemiteTemps	time	56.0812	56.22	0.25%	✅
PeytonManning	MAE_val	0.58159	0.58162	0.0%	✅
PeytonManning	RMSE_val	0.72216	0.72218	0.0%	✅
PeytonManning	Loss_val	0.01239	0.01239	0.01%	✅
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	12.3043	12.12	-1.5%	✅
AirPassengers	MAE_val	13.0626	13.0627	0.0%	✅
AirPassengers	RMSE_val	15.9453	15.9453	0.0%	✅
AirPassengers	Loss_val	0.00131	0.00131	0.0%	✅
AirPassengers	MAE	9.88156	9.88153	-0.0%	✅
AirPassengers	RMSE	11.7354	11.7354	-0.0%	✅
AirPassengers	Loss	0.00052	0.00052	-0.0%	✅
AirPassengers	time	5.18	5.12	-1.16%	✅

Model training plots

Model Training

PeytonManning

YosemiteTemps

AirPassengers

[codecov]

Codecov Report

Merging #1402 (b3b5b36) into main (d236471) will increase coverage by 0.02%.
The diff coverage is 92.30%.

@@            Coverage Diff             @@
##             main    #1402      +/-   ##
==========================================
+ Coverage   89.85%   89.87%   +0.02%     
==========================================
  Files          38       38              
  Lines        5068     5069       +1     
==========================================
+ Hits         4554     4556       +2     
+ Misses        514      513       -1     

Files Changed	Coverage Δ
neuralprophet/time_dataset.py	95.58% <85.71%> (+0.04%)	⬆️
neuralprophet/data/process.py	93.88% <100.00%> (ø)
neuralprophet/df_utils.py	94.65% <100.00%> (+0.14%)	⬆️

[leoniewgnr]

This line is unnecessary as they will be converted to datetime the line below

[noxan]

LGTM