This is not urgent, but useful when a programmer needs a new (small) task.
A number of dialogues make use of the set of distributions. They are in particular the
Model > Probability Distributions > Show Model and also
Model > Probability Distributions > Random Samples and
Model > One Variable > General
a) There are 4 distributions missing from the stats package and these could usefully be added and checked: This could be done together or any one added would be useful:
a1) Beta distribution
a2) Cauchy Distribution
a3) Hypergeometric distribution
a4) Multinomial distribution
b) The birthday distribution is useful. First check (script window by tweaking from another distribution) that it is useful for the show model dialogue. Then it is different, because it is not relevant for the random samples or the fit model dialogues.
c) Similarly check if arima.sim can be added (just called arima) This is to simulate from an arima process. It is for random samples, but not for the probability distributions or the fitting.
d) We already have chi-square and t distributions. But they each could have a non-centrality parameter. This should be easy to add. It applies generally to the show model, random sampling and fitting.
This is not urgent, but useful when a programmer needs a new (small) task. A number of dialogues make use of the set of distributions. They are in particular the Model > Probability Distributions > Show Model and also Model > Probability Distributions > Random Samples and Model > One Variable > General
a) There are 4 distributions missing from the stats package and these could usefully be added and checked: This could be done together or any one added would be useful: a1) Beta distribution a2) Cauchy Distribution a3) Hypergeometric distribution a4) Multinomial distribution
b) The birthday distribution is useful. First check (script window by tweaking from another distribution) that it is useful for the show model dialogue. Then it is different, because it is not relevant for the random samples or the fit model dialogues.
c) Similarly check if arima.sim can be added (just called arima) This is to simulate from an arima process. It is for random samples, but not for the probability distributions or the fitting.
d) We already have chi-square and t distributions. But they each could have a non-centrality parameter. This should be easy to add. It applies generally to the show model, random sampling and fitting.