Installation instructions

From Github:

The github version is usually several weeks ahead of the version on CRAN. You can install it from the Github repository in three different ways.

  1. Run the following in an R session:

    On Windows and MacOSX systems, this will cause a precompiled version of the latest release of pomp to be installed.

  2. Use devtools:
  3. Download the source tarball of the latest release and install it locally as you would any R package.

From CRAN:

Source and binaries for the CRAN version are available here.

Changes in version 2:

pomp version 2 is not fully backward compatible with earlier versions. This means that code that ran under versions <2 may break. An upgrade guide is available to help you transition your codes to the new version.

Important note for Windows users

To make use of pomp’s facilities for accelerated computation using compiled C code, and to compile the package from source, you will need the ability to compile C code and dynamically link it into an R session. For this reason, you must install the Rtools suite, which can be downloaded from Rtools is needed both to compile and install the development version from source and to obtain full value from any version of the package.

When installing Rtools, it is sufficient to choose the “Package authoring installation” option. Also during the installation, tick the “edit system PATH” box.

Important note for Mac OS X users

To make use of the package facilities for accelerated computation using compiled C code, you will need the ability to compile C code and dynamically link it into an R session. These facilities are provided in the Xcode app, which is free and can be installed via the App Store or downloaded from

Some users report problems installing pomp from source due to lack of an appropriate gfortran installation, which is not included by default in all versions of Xcode. If you have this problem, see these instructions.


This software has been made possible by support from the U.S. National Science Foundation (Grants #EF-0545276, #EF-0430120), by the “Inference for Mechanistic Models” Working Group supported by the National Center for Ecological Analysis and Synthesis (a Center funded by N.S.F. (Grant #DEB-0553768), the University of California, Santa Barbara, and the State of California), and by the RAPIDD program of the Science & Technology Directorate, Department of Homeland Security and the Fogarty International Center, U.S. National Institutes of Health.