Skip to content

A comparison between two ways of doing a time series forecast, one using statsmodels + pmdarima packages and the other one using Facebook prophet.

Notifications You must be signed in to change notification settings

homocuadratus/time-series-final-work

Repository files navigation

Time series final work

Objective

To learn how to use Python's statsmodel, pmdarima and fbprophet Packages to analyze and forecast time series.

How to run the notebooks

Please, select a folder of your interest and download this repository to it. Remember that Anaconda python distribution should be installed prior to the execution of any content. To install it, please follow the instructions here:

https://docs.anaconda.com/anaconda/install/index.html

Once you have downloaded this repository, go inside the time-series-final-work folder and execute the following command to create an ad-hoc environment to work:

conda env create -f environment.yml --prefix ./time-series-final-work-env

After running this command, you will see a message saying that, in order to activate this new environment, you should execute

conda activate full/path/to/time-series-final-work-env

Please, do as requested.

Finally, in order to be able to access this environment from within a Jupyter Lab Session execute the following command,

python -m ipykernel install --user --name time-series-final-work-env --display-name="time-series-final-work-ker"

followed by

jupyter lab

and select the notebook of interest.

About

A comparison between two ways of doing a time series forecast, one using statsmodels + pmdarima packages and the other one using Facebook prophet.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published