This is a playground to understand time series and implement related algorithms. Most implemented algorithms are taken from Time Series: Theory and Methods by Brockwell and Davis and mathematical notation is close. One notebook contains a quick summary of a few important points from some chapters and some solutions to the exercises. The other notebook contains examples of some implemented algorithms, these include:

• Yule Walker equations to fit AR models
• Durbin Levinson recursions, maximal likelihood (both via innovation algorithm and Kalman recursions) and least squaress to fit AR, MA and ARMA models
• confidence intervals for parameter estimates by Yule Walker and Durbin Levinson estimates
• various tests for model order selection (AICC, AIC, BIC, FPE)
• goodness of fit tests (turning point test, sign difference test)
• impulse response function
• transformations (mean adjustment, box-cox)
• theoretical and sample acf and pacf

Most functions are unittested.