@ -1527,7 +1527,7 @@ class FitDistribution(rv_frozen):
'P-value is on the conservative side (i.e. too large) due to' +
' ties in the data!')
T = self.dist.nlogps(theta, x)
T = self._nlogps(theta, x)
n = len(x)
np1 = n + 1