Alexey Izmailov
Alexey Izmailov
The [current implementation](https://mc-stan.org/docs/functions-reference/wiener-first-passage-time-distribution.html) of the Wiener First Passage Time Distribution lacks a generation step for a variate following this distribution. This feature should be analogous to the ``_rng`` methods that...
## Description Analogous to other distributions, add support for the [inverse Gaussian distribution](https://www.wikiwand.com/en/Inverse_Gaussian_distribution) with a sampling statement and Stan functions for working with its log density, CDF, and CCDF. This...
Now that additional models are being added, lots of code is being reused. Method arguments, constants, and interfacing should all be improved for cleanliness and modularity.
Analogous to the Beta regression given in rstanarm [here](http://mc-stan.org/rstanarm/reference/stan_betareg.html), have the same feature set. This should be an independent class from the existing GLM as otherwise the API will be...
This model lends itself well to the ``fit()`` and ``predict()`` methodology. For a simple example, see [this](http://singmann.org/wiener-model-analysis-with-brms-part-i/) and its [sequel](http://singmann.org/wiener-model-analysis-with-brms-part-ii/). Regression should occur on every model parameter specified in the...
The package needs an example of adding models and GAM would be a natural next step. See [here](https://github.com/stan-dev/rstanarm/blob/master/R/stan_gamm4.R) and [here](https://github.com/stan-dev/example-models/tree/master/misc/gam) for implementation examples.
It would be nice to illustrate the control over priors that users have - alpha, beta, and auxiliary parameters can all be set to different distributions with chosen parameters. Additionally,...
GPRs lend themselves well to the ``fit()``, ``predict()`` methodology. Moreover, a well-done GPR will enable this package to be used for kriging, which has been used for the [solution of...