Interview Assignment 1 - Sep2023
“Decomposition of time series data”
The data csv file can be found HERE.
Using Jupyter Notebook, apply a time series decomposition model to the data to identify the components of trend, cycle and noise.
A good resource you might like to refer to is here: https://otexts.com/fpp3/
Include a plot similar to Figure 3.13 in this webpage: https://otexts.com/fpp3/classical-decomposition.html
You may consider STL or Seasonal ARIMA or any others you wish.
The data granularity is hourly. The seasonality is weekly. Use your discretion regarding the trend evolution/window, but we suggest checking 7 or 30 or 90 days.
Please include some brief notes on your choice of model and any observations or ideas you see in the data.
Please email us the Notebook within a ZIP file to [email protected] or via www.wetransfer.com (as often code attachments get blocked by email servers).
The data csv file can be found HERE.
Using Jupyter Notebook, apply a time series decomposition model to the data to identify the components of trend, cycle and noise.
A good resource you might like to refer to is here: https://otexts.com/fpp3/
Include a plot similar to Figure 3.13 in this webpage: https://otexts.com/fpp3/classical-decomposition.html
You may consider STL or Seasonal ARIMA or any others you wish.
The data granularity is hourly. The seasonality is weekly. Use your discretion regarding the trend evolution/window, but we suggest checking 7 or 30 or 90 days.
Please include some brief notes on your choice of model and any observations or ideas you see in the data.
Please email us the Notebook within a ZIP file to [email protected] or via www.wetransfer.com (as often code attachments get blocked by email servers).