python - How to convolve two distirbutions from scipy library -

python - How to convolve two distirbutions from scipy library -

i have seen (by researching) convolution beingness done via numpy, if wish convolve 2 standard distributions (specifically normal uniform) readily available in scipy library, there direct way of doing rather creating 2 arrays via numpy , convolve?

in general, computing convolutions distributions requires solving integrals. worked on problem part of work dissertation [1] , wrote java (rather idiosyncratic) carry out operations. approach create catalog of distributions there known results, , fall on numerical method (convolution via discretization , fft) if there no known result.

for combination of gaussian , uniform, result gaussian bump split 2 , pasted onto each end of uniform distribution, when uniform wide enough, otherwise looks bump. can seek find formulas if interested.

you can seek compute integrals via symbolic computation scheme such maxima. [2] example, maxima says convolution of unit gaussian unit uniform is:

-(erf((sqrt(2)*s-sqrt(2))/2)-erf(s/sqrt(2)))/2

[1] http://riso.sourceforge.net/docs/dodier-dissertation.pdf (in particular section c.3.17)

[2] http://sourceforge.net/p/maxima

python statistics scipy