london
Python analysis of London housing prices from 1995 to 2016
Code summary (nb-UK.ipynb):
- Univariate linear regression
- Univariate exponential regression
- Univariate logreg
- Holt-Winter
- Multivariate regression (see doc.pdf for included factors)
- Trends decomposition and seasonal factor computation
- ACF and PACF
- stationnarity induction : ordinal difference, saisonal difference, combined, 2nd order combined
- Arima model (autotraining, on 3 datasets : residual from the trend decomposition - 2nd order+saisonal difference - automated sarima)
All the code is avalaibla as a Jupyter notebook (nb-UK.ipynb). Detailed analysis avalaible in French only (analyse.pdf), if any translation needed, feel free to ask me
- some part of the analysis done with Time Series Expert (see www.itse.be)