search

AutoViML / Auto_TS

Automatically build ARIMA, SARIMAX, VAR, FB Prophet and XGBoost Models on Time Series data sets with a Single Line of Code. Created by Ram Seshadri. Collaborators welcome.
Apache License 2.0
721 stars 113 forks source link

'DataFrame' object has no attribute 'append' #113

Open adityamalikzeiss opened 1 month ago

adityamalikzeiss commented 1 month ago

I am getting the following exceptions while trying Auto_TS:

Exception occurred while building Prophet model... 'DataFrame' object has no attribute 'append' FB Prophet may not be installed or Model is not running...

Exception occurred while building Auto SARIMAX model... 'DataFrame' object has no attribute 'append' Auto SARIMAX model error: predictions not available.

I have followed the example notebook given for univariate model.

adityamalikzeiss commented 1 month ago

Getting the exact same errors when executing the autots_univariate_example.ipynb:

Best model is a Seasonal SARIMAX(2,1,1)*(0,0,0,12), aic = 2458.432 Exception occurred while building Auto SARIMAX model... 'DataFrame' object has no attribute 'append' Auto SARIMAX model error: predictions not available.

Fold Number: 2 --> Train Shape: 441 Test Shape: 5 Exception occurred while building Prophet model... 'DataFrame' object has no attribute 'append' FB Prophet may not be installed or Model is not running...

BryR0 commented 1 month ago

replace line models\build_*.py: extra_concatenated = extra_concatenated.append(concatenated)

to `**try:

extra_concatenated = extra_concatenated.append(concatenated)

except Exception as e:

extra_concatenated = pd.concat([extra_concatenated,concatenated])** `

in line 968 auto_ts__init__.py:

Now make predictions using model given

predictions=None <---- local variable