search

timeseriesAI / tsai

Time series Timeseries Deep Learning Machine Learning Python Pytorch fastai | State-of-the-art Deep Learning library for Time Series and Sequences in Pytorch / fastai
https://timeseriesai.github.io/tsai/
Apache License 2.0
5.16k stars 644 forks source link

get_Monash_data(dsid) should be get_Monash_regression_data(dsid) #146

Closed meanpenguin closed 3 years ago

meanpenguin commented 3 years ago

In the tutorial_nbs/10_Time_Series_Classification_and_Regression_with_MiniRocket.ipynb notebook errors

NameError                                 Traceback (most recent call last)
<ipython-input-11-004aad2fdf41> in <module>()
      2 from sklearn.metrics import mean_squared_error
      3 dsid = 'AppliancesEnergy'
----> 4 X_train, y_train, X_valid, y_valid = get_Monash_data(dsid)
      5 rmse_scorer = make_scorer(mean_squared_error, greater_is_better=False)
      6 model = MiniRocketRegressor(scoring=rmse_scorer)

NameError: name 'get_Monash_data' is not defined

Looks like a function was renamed: get_Monash_data(dsid) should now be get_Monash_regression_data(dsid)

# Univariate regression with sklearn-type API
from sklearn.metrics import mean_squared_error
dsid = 'Covid3Month'
X_train, y_train, X_valid, y_valid = get_Monash_data(dsid)
fastai     : 2.4
fastcore   : 1.3.20
torch      : 1.8.1
tsai       : 0.2.18
oguiza commented 3 years ago

Thanks @meanpenguin for letting me know. I've already updated this in the nb.

meanpenguin commented 3 years ago

Confirmed fix.