BUG: `laplace.inverse_cdf()` is not correct.

[Warren Weckesser]

The implementation of the inverse_cdf() method for the Laplace distribution is missing calls to ln().
The following

statrs/src/distribution/laplace.rs

Lines 127 to 131 in 435739b

 if p <= 0.5 { 
     self.location + self.scale * (2. * p) 
 } else { 
     self.location - self.scale * (2. - 2. * p) 
 } 

should be

        if p <= 0.5 {
            self.location + self.scale * (2. * p).ln()
        } else {
            self.location - self.scale * (2. - 2. * p).ln()
        }

(See, for example, the entry for Quantile in the table on the right in the Laplace distribution wikipedia page.)

I'll submit a pull request with a fix.