01_getting_started_complete

[kgboyko]

Description

Description

Block:

nf = NeuralForecast(
models=[
AutoNHITS(h=48, config=config_nhits, loss=MQLoss(), num_samples=5),
AutoLSTM(h=48, config=config_lstm, loss=MQLoss(), num_samples=2),
],
#freq='H',
)
nf.fit(df=Y_df)

Required replace freq='H' on freq=1

There is a tax block in another place, but with cross_validation:

nf = NeuralForecast(
models=[
AutoNHITS(h=48, config=config_nhits, loss=MQLoss(), num_samples=5),
AutoLSTM(h=48, config=config_lstm, loss=MQLoss(), num_samples=2),
],
freq='H',
)
cv_df = nf.cross_validation(Y_df, n_windows=2)

Likewise replace on freq=1.

---

ValueError Traceback (most recent call last)
Cell In[39], line 11
1 #%%capture
2 nf1 = NeuralForecast(
3 models=[
4 AutoNHITS(h=48, config=config_nhits, loss=MQLoss(), num_samples=5),
(...)
8 #freq=1, # https://pandas-pydata-org.translate.goog/pandas-docs/stable/user_guide/timeseries.html?_x_tr_sl=en&_x_tr_tl=ru&_x_tr_hl=ru&_x_tr_pto=sc#offset-aliases
9 )
---> 11 nf1.fit(df=Y_df)

File /opt/conda/lib/python3.10/site-packages/neuralforecast/core.py:495, in NeuralForecast.fit(self, df, static_df, val_size, sort_df, use_init_models, verbose, id_col, time_col, target_col, distributed_config, prediction_intervals)
493 # Process and save new dataset (in self)
494 if isinstance(df, (pd.DataFrame, pl_DataFrame)):
--> 495 validate_freq(df[time_col], self.freq)
496 self.dataset, self.uids, self.last_dates, self.ds = self._prepare_fit(
497 df=df,
498 static_df=static_df,
(...)
503 target_col=target_col,
504 )
505 self.sort_df = sort_df

File /opt/conda/lib/python3.10/site-packages/utilsforecast/validation.py:140, in validate_freq(times, freq)
135 def validate_freq(
136 times: Series,
137 freq: Union[str, int],
138 ) -> None:
139 if _is_int_dtype(times) and not isinstance(freq, int):
--> 140 raise ValueError(
141 "Time column contains integers but the specified frequency is not an integer. "
142 "Please provide a valid integer, e.g. freq=1"
143 )
144 if _is_dt_dtype(times) and isinstance(freq, int):
145 raise ValueError(
146 "Time column contains timestamps but the specified frequency is an integer. "
147 "Please provide a valid pandas or polars offset, e.g. freq='D' or freq='1d'."
148 )

ValueError: Time column contains integers but the specified frequency is not an integer. Please provide a valid integer, e.g. freq=1

Link

No response

Link

No response

[jmoralez]

Fixed by 2.0 release