Adapt methods

[mnlang]

Hello Alex,

according to our communication in PR #67, I have renamed the arguments from d to x for the functions computing the moments of probability distributions in order to be in line with the generic mean(). I have adapted this equivalently for random() and quantile(), allowing a regular dispatch to the generic quantile().

Summary of changes:

• Renaming of arguments d to x in methods mean(), variance(), skewness(), kurtosis(), quantile().
• Dispatch all quantile methods to the generic stats::quantile().
• Adapt manual for variance which now refers to the generics mean() and variance().
• Additionally, I've added checks if ... argument is evaluated within mean methods.

Even though there is no distributions3 specific manual page for quantile() anymore, I think this improves the package now, because: 1) This solves the warning regarding stats:::quantile.default and still solves Issue #58. 2) Various packages provide their own quantile methods - these now work with the adpations.

I hope this is in your interest, otherwise I can of course change parts according to your suggestions.

Cheers

[alexpghayes]

Thank for this, this is incredible!