Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Pipeline does not work when using FourierFeaturizer in colab #540

Open
ams015 opened this issue Mar 7, 2023 · 1 comment
Open

Pipeline does not work when using FourierFeaturizer in colab #540

ams015 opened this issue Mar 7, 2023 · 1 comment
Labels
🪲 : bug

Comments

@ams015
Copy link

ams015 commented Mar 7, 2023

Describe the bug

I have run this code locally and it has worked, but when I run it on colab I face a value error

To Reproduce

import pandas as pd
from pmdarima import pipeline
from pmdarima import model_selection
from pmdarima import preprocessing as ppc
from pmdarima import arima

#my dataset has 2k points, but I am going to give the first 10 points for reproducibilit
A=[81.99130735,
82.02854792,
82.02854792,
79.41550171,
80.50789157,
64.73651292,
70.85637886,
75.05835578,
75.89006169,
72.05428365]
B=pd.DatetimeIndex(['2013-01-01', '2013-01-02', '2013-01-03', '2013-01-04',
'2013-01-05', '2013-01-06', '2013-01-07', '2013-01-08',
'2013-01-09', '2013-01-10'],
dtype='datetime64[ns]', name='Date', freq='D')

xt_a_train=pd.DataFrame(A,index=B)
pipe = pipeline.Pipeline([
("fourier", ppc.FourierFeaturizer(m=365.25,k=4)),
("arima", arima.AutoARIMA(stepwise=True, trace=1, error_action="ignore",start_p=4, start_q=0,d=0,
seasonal=True,m=7 ,
suppress_warnings=True))])
pipe.fit(xt_a_train)
print("Model fit:")

Versions

System:
    python: 3.8.10 (default, Nov 14 2022, 12:59:47)  [GCC 9.4.0]
executable: /usr/bin/python3
   machine: Linux-5.10.147+-x86_64-with-glibc2.29

Python dependencies:
        pip: 22.0.4
 setuptools: 57.4.0
    sklearn: 1.2.1
statsmodels: 0.13.5
      numpy: 1.22.4
      scipy: 1.10.1
     Cython: 0.29.33
     pandas: 1.3.5
     joblib: 1.2.0
   pmdarima: 2.0.2
Linux-5.10.147+-x86_64-with-glibc2.29
Python 3.8.10 (default, Nov 14 2022, 12:59:47) 
[GCC 9.4.0]
pmdarima 2.0.2
NumPy 1.22.4
SciPy 1.10.1
Scikit-Learn 1.2.1
Statsmodels 0.13.5

Expected Behavior

Code should run correctly and pick the correct model; it does so on desktop.

Actual Behavior

Error thrown
Performing stepwise search to minimize aic
ARIMA(4,0,0)(1,0,1)[7] intercept : AIC=inf, Time=nan sec
ARIMA(0,0,0)(0,0,0)[7] intercept : AIC=inf, Time=nan sec
ARIMA(1,0,0)(1,0,0)[7] intercept : AIC=inf, Time=nan sec
ARIMA(0,0,1)(0,0,1)[7] intercept : AIC=inf, Time=nan sec
ARIMA(0,0,0)(0,0,0)[7] : AIC=inf, Time=nan sec
ARIMA(4,0,0)(0,0,1)[7] intercept : AIC=inf, Time=nan sec
ARIMA(4,0,0)(1,0,0)[7] intercept : AIC=inf, Time=nan sec
ARIMA(4,0,0)(2,0,1)[7] intercept : AIC=inf, Time=nan sec
ARIMA(4,0,0)(1,0,2)[7] intercept : AIC=inf, Time=nan sec
ARIMA(4,0,0)(0,0,0)[7] intercept : AIC=inf, Time=nan sec
ARIMA(4,0,0)(0,0,2)[7] intercept : AIC=inf, Time=nan sec
ARIMA(4,0,0)(2,0,0)[7] intercept : AIC=inf, Time=nan sec
ARIMA(4,0,0)(2,0,2)[7] intercept : AIC=inf, Time=nan sec
ARIMA(3,0,0)(1,0,1)[7] intercept : AIC=inf, Time=nan sec
ARIMA(5,0,0)(1,0,1)[7] intercept : AIC=inf, Time=nan sec
ARIMA(4,0,1)(1,0,1)[7] intercept : AIC=inf, Time=nan sec
ARIMA(3,0,1)(1,0,1)[7] intercept : AIC=inf, Time=nan sec
ARIMA(5,0,1)(1,0,1)[7] intercept : AIC=inf, Time=nan sec
ARIMA(4,0,0)(1,0,1)[7] : AIC=inf, Time=nan sec

ValueError Traceback (most recent call last)
in
----> 1 pipe.fit(xt_a_train)
2 print("Model fit:")
3 print(pipe)

4 frames
/usr/local/lib/python3.8/dist-packages/pmdarima/pipeline.py in fit(self, y, X, **fit_kwargs)
220 # Now fit the final estimator
221 kwargs = named_kwargs[steps[-1][0]]
--> 222 self._final_estimator.fit(yt, X=Xt, **kwargs)
223 return self
224

/usr/local/lib/python3.8/dist-packages/pmdarima/arima/auto.py in fit(self, y, X, **fit_args)
165 sarimax_kwargs = {} if not self.kwargs else self.kwargs
166
--> 167 self.model_ = auto_arima(
168 y,
169 X=X,

/usr/local/lib/python3.8/dist-packages/pmdarima/arima/auto.py in auto_arima(y, X, start_p, d, start_q, max_p, max_d, max_q, start_P, D, start_Q, max_P, max_D, max_Q, max_order, m, seasonal, stationary, information_criterion, alpha, test, seasonal_test, stepwise, n_jobs, start_params, trend, method, maxiter, offset_test_args, seasonal_test_args, suppress_warnings, error_action, trace, random, random_state, n_fits, return_valid_fits, out_of_sample_size, scoring, scoring_args, with_intercept, sarimax_kwargs, **fit_args)
699 )
700
--> 701 sorted_res = search.solve()
702 return _return_wrapper(sorted_res, return_valid_fits, start, trace)
703

/usr/local/lib/python3.8/dist-packages/pmdarima/arima/_auto_solvers.py in solve(self)
458 )
459
--> 460 sorted_fits = _sort_and_filter_fits(filtered_models_ics)
461 if self.trace and sorted_fits:
462 print(f"\nBest model: {str(sorted_fits[0])}")

/usr/local/lib/python3.8/dist-packages/pmdarima/arima/_auto_solvers.py in _sort_and_filter_fits(models)
566 # if the list is empty, or if it was an ARIMA and it's None
567 if not filtered:
--> 568 raise ValueError(
569 "Could not successfully fit a viable ARIMA model "
570 "to input data.\nSee "

ValueError: Could not successfully fit a viable ARIMA model to input data.
See http://alkaline-ml.com/pmdarima/no-successful-model.html for more information on why this can happen.

Additional Context

No response
@aaronreidsmith
Copy link
Member

I am unable to reproduce either locally or in Google Colab (link). Does your example input work locally? I get this in both environments:

Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "/Users/asmith/.pyenv/versions/3.10.6/lib/python3.10/site-packages/pmdarima/pipeline.py", line 222, in fit
    self._final_estimator.fit(yt, X=Xt, **kwargs)
  File "/Users/asmith/.pyenv/versions/3.10.6/lib/python3.10/site-packages/pmdarima/arima/auto.py", line 167, in fit
    self.model_ = auto_arima(
  File "/Users/asmith/.pyenv/versions/3.10.6/lib/python3.10/site-packages/pmdarima/arima/auto.py", line 506, in auto_arima
    D = nsdiffs(xx, m=m, test=seasonal_test, max_D=max_D,
  File "/Users/asmith/.pyenv/versions/3.10.6/lib/python3.10/site-packages/pmdarima/arima/utils.py", line 106, in nsdiffs
    dodiff = testfunc(x)
  File "/Users/asmith/.pyenv/versions/3.10.6/lib/python3.10/site-packages/pmdarima/arima/seasonality.py", line 597, in estimate_seasonal_differencing_term
    stat = self._compute_test_statistic(x)
  File "/Users/asmith/.pyenv/versions/3.10.6/lib/python3.10/site-packages/pmdarima/arima/seasonality.py", line 537, in _compute_test_statistic
    fit = self._fit_ocsb(x, m, lag_term, maxlag)
  File "/Users/asmith/.pyenv/versions/3.10.6/lib/python3.10/site-packages/pmdarima/arima/seasonality.py", line 493, in _fit_ocsb
    ar_fit = sm.OLS(y, add_constant(mf)).fit(method='qr')
  File "/Users/asmith/.pyenv/versions/3.10.6/lib/python3.10/site-packages/statsmodels/tools/tools.py", line 270, in add_constant
    is_nonzero_const = np.ptp(x, axis=0) == 0
  File "<__array_function__ internals>", line 180, in ptp
  File "/Users/asmith/.pyenv/versions/3.10.6/lib/python3.10/site-packages/numpy/core/fromnumeric.py", line 2669, in ptp
    return _methods._ptp(a, axis=axis, out=out, **kwargs)
  File "/Users/asmith/.pyenv/versions/3.10.6/lib/python3.10/site-packages/numpy/core/_methods.py", line 279, in _ptp
    umr_maximum(a, axis, None, out, keepdims),
ValueError: zero-size array to reduction operation maximum which has no identity

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees
No one assigned
Labels
🪲 : bug
Projects
None yet
Milestone
No milestone
Development

No branches or pull requests
2 participants
@ams015 @aaronreidsmith