README.md

In hunzikp/spmle: Binary Choice Models Using PMLE

spmle

Philipp Hunziker June 21, 2020

System Requirements

To use the (optional but recommended) Python speed-up, you need Python 3.6+ (and conda if on Windows).

Installation

Install spmle directly from this repository:

devtools::install_github("hunzikp/spmle")

In case you don’t have scipy and numpy installed (or don’t know what these are) and you want to use the Python speed-up, run the following in an R console:

spmle::install_pydep()

This command installs the Python dependencies in the r-reticulate virtual environment.

Usage

To use the Python speed-up you need to specify a python binary or virtual environment where scipy and numpy are installed. If you ran the install_pydep() command above, then you can use the following:

library(reticulate)
library(spmle)
reticulate::use_virtualenv("r-reticulate")  # In Unix
reticulate::use_condaenv("r-reticulate")  # In Windows

Examples

See here.

hunzikp/spmle documentation built on June 28, 2020, 10:30 a.m.