./gen-data/:
C++ code that procedurally generates time-series data according to a specified stochastic process. Data is output as CSVs. The following processes are supported:
- Autoregressive (AR) processes
- Autoregressive processes with polynomial dependence on lagged values
- Autoregressive–moving-average (ARMA) processes
- Autoregressive–moving-average processes with exogenous inputs (ARMAX)
./aug-data/:
C++ code that augments existing time-series with additive noise sampled from a Gaussian white noise process. Data is handled as CSVs for both input and output.
This code was originally written when I was doing research at UMD. Although this repo primarily exists as a record of my work, I'll likely edit the READMEs for legibility soon. After all, what good is showcasing code that can't be understood?