Wrong answers from integralBinomialPDF for p = 0.0 and p = 1.0

[gaurav-arya]

For example, integralBinomialPDF 3 1 3 = 2.718281828459045. The issue seems to be that the edge-case handling in integralBinomialLogPdf outputs the actual PDF rather than the log PDF:

random-fu/random-fu/src/Data/Random/Distribution/Binomial.hs

Lines 75 to 80 in b7f296b

	| p == 0.0 && xI == 0 = 1.0
	| p == 0.0 = 0.0
	| p == 1.0 && xI == nI = 1.0
	| p == 1.0 = 0.0
	| xI == 0 = n * log (1-p)
	| xI == nI = n * log p

.