search

susanli2016 / Machine-Learning-with-Python

Python code for common Machine Learning Algorithms
4.31k stars 4.82k forks source link

Make Time Series Forecastings.ipynb work with python 3.10.10 #47

Open gabriel1628 opened 1 year ago

gabriel1628 commented 1 year ago

Make the program work with matplotlib 3.7.0, numpy 1.24.2, pandas 1.5.3 statsmodels 0.13.5 and prophet 1.1.2

gabriel1628 commented 1 year ago

Hello Susan, I've read your tutorial 'An End-to-End Project on Time Series Analysis and Forecasting with Python' on towardsdatascience.com and I really appreciated so I would like to thank you for your work. I had to make some changes in the code to make it work, I guess it's because the librairies have evolved since then. I am working on python 3.10.10 with matplotlib 3.7.0, numpy 1.24.2, pandas 1.5.3 statsmodels 0.13.5 and prophet 1.1.2. Here are some modifications I've brought :

The best results I get with ARIMA are obtained for the parameters (0, 1, 1) and (0, 1, 1, 12) with an AIC of 279.58. However, the MSE and RMSE I get are larger.

When fitting with the ARIMA model, I had to remove enforce_stationarity=False for making the command results.plot_diagnostics(figsize=(16, 8)) work.

fbprophet package is know prophet