Fix API Irregularities Between dstats.distrib and dstats.random

[jmh530]

I have noticed some irregularities in the API for dstats.distrib and dstats.random.

1. Most of the inverse CDF functions are f(p, param), but some are f(param, p). These are the two that jump out as being opposite of the others: invChiSquareCDFR, invFisherCDFR

2. The names in dstats.random do not match the names in dstats.distrib in some cases. These three jump out: rLogNorm, rNorm, rStudentT

3. Excluding the first time (which obviously should be different), the type signature of invBinomialCDF does not match that of binomialCDF. Contrasting invBinomialCDF and binomialCDF, n is uint in one and ulong in another.

[jmh530]

4. Hypergeometric in dstats.distrib are long, but are int in dstats.random. There is a similar issue with rBinomial.