Practical TimeSeries Analysis course on Coursera has all its assignments and notebooks in R. Though I use R sometimes but python is the go-to language for me. So, the project tries to rewrite R notebooks and all the analysis that has explained step and step in python. Datasets are same and have been accumulated in the data folder. The projects describes the usage of:
SARIMA from statsmodels package for experimentation with ARIMA and seasonal ARIMA. This is somewhat equivalent to arima from forecast package in R. pmdarima package for running something similar to auto.arima in R. I tried to find auto.arima in statsmodel but couldn't. Yule Walker Equations step by step using python packages. Time-series visualization functions in statsmodels and seaborn
Next steps: Extending this repo of all possible packages and visualization for time series so that it becomes easy to find the right tool from the enormous list of tools.
Next addition:
Modelling:
- ARIMAX
- Holt-winters
Visz :
- plotly