Fix MGF of Negative Binomial

This release fixes a typo in the MGF of the Negative Binomial distribution. (#29)

Thanks to @acha94 for reporting!

Fix Quotient of Random Variables

The formula that describes a quotient transformation of two random variables contained a mistake. Spotted and fixed by @Rodvi via #20.

Fix CDF of F distribution

The CDF formula of the F distribution contained a typo where d_1 got repeated twice instead of being once d_1 and once d_2.

Spotted by Catherine Tousey.

Fix expectation relationship

This release resolves #19, an invalid equivalence relationship that should have been an implication.

Spotted by @alphaville.

Fix issues in expectation section

This commit fixes two issues reported by @allipatev (#18):

1. A botched formula involving conditional expectation.
2. A missing note that an identity only works for discrete random variables.

Fix variance of sum of correlated variables

This release fixes a typo (Y -> X) in the variance of a sum of correlated random variables.

Contributed by Adrian Bartnik (@whoww)

Fix variance of Pareto distribution

This release fixes a typo in the variance of the Pareto distribution.

Spotted by Carlo Cavalieri.

Fix multinomial expectation and variance

This release primarily fixes how expectation and variance was shown in the multinomial distribution (#14). In particular, @faridcher contributed the following changes:

• Use matrix instead of scalar form for multinomial (b0a8912)
• Exclude *.gz files created during compilation (3b30e26)
• Add missing d\theta in posterior mean (af30a34)
• Add constraint on Students't expectation (7d0082e)

This release also improves the build process on Windows (#15).

Clarify Parameterization of Exponential and Gamma Distribution

This release includes the following changes:

• Fix inconsistency between parameterization of Exponential and Gamma distribution (#13).
• Add new special cases of the Gamma function.
• Switch table-of-contents color from aggressive red to black.
• Use lighter blue for link and citation colors.

Corrected Exponential and Gamma Plots

The parameters in the plots from exponential and gamma versions did not correspond to what the plots were showing. While the legend contained the scale parameter beta, the plot parameterization used the rate parameter lambda.

Caught by Farid Cheraghi.