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.

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Google Colab import error

nicolasseverino opened this issue · comments

Hi ! I wanted to try out your framework but although I manage to install it on Google Colab, I can't import it :

If I write in a cell :

from auto_ts import auto_timeseries

I get :

"load() missing 1 required positional argument: 'Loader'"

Obviously, it seems coming from "yaml.load(f)". How can I import and use it on Google Colab then please ?

Thanks for your help !

Hi @nicolasseverino 👍
Please double check if you are using the right "auto_ts" library. There are 2 libraries with similar name: "autots" and "auto_ts".

Second, please paste a screenshot of the error. I could have seen where the error was coming from. Otherwise, it is hard.

Thanks
AutoViML

Hi @nicolasseverino :

The following three steps worked for me in Colab - do this one after another in each cell.

!pip install auto-ts --no-deps --ignore-installed
!pip install 'fsspec>=0.3.3'
!pip install statsmodels

But please restart the runtime (look up in the Colab Runtime menu) after installing statsmodels and then do this:

from auto_ts import auto_timeseries

Voila! that worked for me.
Autovimal

Still isn't working at all for me. Same error, tried all the installations also with Conda on Colab execution fails. Is there any alternative for cloud services?

from auto_ts import auto_timeseries

Error following exact steps:


ModuleNotFoundError Traceback (most recent call last)

in
----> 1 from auto_ts import auto_timeseries

3 frames

/usr/local/lib/python3.7/dist-packages/auto_ts/models/ar_based/build_autoarimax.py in
6 # TODO: Resolve which one we want to use
7 #from pmdarima.arima.auto import auto_arima # type: ignore
----> 8 from pmdarima.arima import auto_arima # type: ignore
9
10

ModuleNotFoundError: No module named 'pmdarima'


NOTE: If your import is failing due to a missing package, you can
manually install dependencies using either !pip or !apt.

To view examples of installing some common dependencies, click the
"Open Examples" button below.

from auto_ts import auto_timeseries

Error :
OSError: dlopen(/Users/name/opt/anaconda3/lib/python3.9/site-packages/lightgbm/lib_lightgbm.so, 6): Library not loaded: /usr/local/opt/libomp/lib/libomp.dylib
Referenced from: /Users/name/opt/anaconda3/lib/python3.9/site-packages/lightgbm/lib_lightgbm.so
Reason: image not found