fix notation in laplace.jl docstring

[JeffFessler]

No description provided.

[matbesancon]

I'm fine to keep these conventions (better than being inconsistent), but a great place of knowledge uses b instead of beta

[JeffFessler]

I am fine with b too - feel free to revert. I think I saw \beta in some places and \theta in others (and b in others) so I was just pursuing self-consistency.
I looked in several probability texts on my shelf and most of them write it in terms of \lambda = 1/b and one book used \alpha=1/b so there is no standard out there it seems. Cheers.

[matbesancon]

Just waiting for another opinion and this is merged, thanks for the contribution!

[matbesancon]

(Travis is failing while #836 not merged)

[jmxpearson]

Looks good to me.

[simonbyrne]

Should it be sigma?

[matbesancon]

doesn't seem to be the convention for Laplace.
Std is sqrt(2)*b

[simonbyrne]

we should be consistent for all location/scale families:

• Cauchy
• Logistic
• SymTriangular
• Biweight
• Triweight
• Epanechnikov
• Cosine

[matbesancon]

This part will be simpler for the construction with #823, but without it, it may be dangerous to require a parameter which is not the one usually taken

[simonbyrne]

I’m no sure what you mean?

[JeffFessler]

I just noticed that cauchy.jl has similar inconsistencies in its docstring.
Perhaps @simonbyrne is suggesting that all docstrings should be made more self consistent.
I agree with the goal of self consistency within each docstring, but I also agree with @matbesancon that trying to unify notation across all distributions could be problematic. For example, I see that the docstring for cauchy.jl uses \mu and \sigma but that is nonstandard and possibly misleading notation because Cauchy does not have a mean or variance.

[matbesancon]

I'd be for keeping letters different from mu, sigma when the location and scale parameters do not represent mean, std directly

[matbesancon]

restarted the build, merging as everything goes green

[matbesancon]

@simonbyrne can you restart appveyor here?