[FIX] Feature importance when using only hist_exog in TFT fails

[elephaint]

Fixes #1169. The below code fails without the fix:

import pandas as pd
import matplotlib.pyplot as plt
import numpy as np
from neuralforecast import NeuralForecast
from neuralforecast.models import TFT
from neuralforecast.losses.pytorch import DistributionLoss
from neuralforecast.utils import AirPassengersPanel, AirPassengersStatic

AirPassengersPanel['month']=AirPassengersPanel.ds.dt.month
Y_train_df = AirPassengersPanel[
    AirPassengersPanel.ds < AirPassengersPanel["ds"].values[-24]
]  # 120 train
Y_test_df = AirPassengersPanel[
    (AirPassengersPanel.ds >= AirPassengersPanel["ds"].values[-24])
    & (AirPassengersPanel.ds < AirPassengersPanel["ds"].values[-12])
].reset_index(
    drop=True
)  # 12 test
nf = NeuralForecast(
    models=[
        TFT(
            h=12,
            input_size=48,
            hidden_size=20,
            loss=DistributionLoss(distribution="StudentT", level=[80, 90]),
            learning_rate=0.005,
            # stat_exog_list=[],
            # futr_exog_list=['trend'],
            hist_exog_list=['trend'],
            max_steps=300,
            val_check_steps=10,
            early_stop_patience_steps=10,
            scaler_type="robust",
            windows_batch_size=None,
            enable_progress_bar=True,
        ),
    ],
    freq="M",
)
nf.fit(df=Y_train_df, static_df=AirPassengersStatic, val_size=12)
Y_hat_df = nf.predict(futr_df=Y_test_df)

# Plot quantile predictions
Y_hat_df = Y_hat_df.reset_index(drop=False).drop(columns=["unique_id", "ds"])
plot_df = pd.concat([Y_test_df, Y_hat_df], axis=1)
plot_df = pd.concat([Y_train_df, plot_df])

plot_df = plot_df[plot_df.unique_id == "Airline1"].drop("unique_id", axis=1)
plt.plot(plot_df["ds"], plot_df["y"], c="black", label="True")
plt.plot(plot_df["ds"], plot_df["TFT"], c="purple", label="mean")
plt.plot(plot_df["ds"], plot_df["TFT-median"], c="blue", label="median")
plt.fill_between(
    x=plot_df["ds"][-12:],
    y1=plot_df["TFT-lo-90"][-12:].values,
    y2=plot_df["TFT-hi-90"][-12:].values,
    alpha=0.4,
    label="level 90",
)
plt.legend()
plt.grid()
plt.plot()

Y_hat_df = nf.predict(futr_df=Y_test_df)
feature_importances = nf.models[0].feature_importances()

[review-notebook-app]

Check out this pull request on 

See visual diffs & provide feedback on Jupyter Notebooks.

---

Powered by ReviewNB

[marcopeix]

LGTM