[Gretl-devel] nonlinearity tests, etc. (2)

On Sun, 28 Feb 2021, Sven Schreiber wrote: I can see this could be useful if you just have a residual series you want to test (e.g. after ARMA). So I've put it into git on a trial basis. There's a new C-plugin, bdstest, which supports (for now) a "hidden" function, _bdstest(). It's based on LeBaron's "fast" code. The signature of the hidden function is: matrix _bdstest(series x, int m, scalar eps) where m is the embedding level (>= 2) and eps the cut-off for "closeness". Right now eps is just a raw number in the units of x, but if we pursue this it should probably be expressed as a fraction of the standard deviation of x (a rule of thumb seems to be eps = 0.5*sd(x)). The return value is a column vector of m-1 z-scores, for testing at embedding levels 2 to m. I'm attaching LeBaron's sample data as x.txt, and the output from his original program as output.txt. You can replicate this via <hansl> open x.txt matrix B = _bdstest(x, 5, 100) print B </hansl> The plugin file bdstest.c can also be copied out of the source tree and compiled as a stand-alone "libgretl program", by defining STANDALONE to 1 near the top of the file. In that mode the program does Monte Carlo, which reveals that the test over-rejects quite substantially for series of typical econometric length -- so if we pursue this we may want to look into MBDS. The MC parameters are mostly hard-wired; edit the C to change things. Allin