Installing¶

Dependencies¶

Requirements

These are external packages which you will need to install before installing Basemap.

Matplotlib 1.0.0 (or later, download)

Python 2.6 (or later, including Python 3) (download): Matplotlib 2.2 LTS requires Python 2.7 or later Matplotlib 3.0 requires Python 3.5 or later
NumPy 1.2.1 (or later): Array support for Python (download)

PROJ4 Cartographic Projections Library.

Required library that ships with Basemap

GEOS (Geometry Engine - Open Source) library 3.1.1 or later.: Source code is included in the geos-3.3.3 directory. When building from source, must be built and installed separately from basemap (see build instructions below). Included in Windows binary installers.

Optional libraries

Pillow: Python Imaging Library (download), only needed for bluemarble(), etopo(), shadedrelief() and warpimage() instance methods.

Download either Windows binary installers or source tarballs here.

To install from the source, follow these steps:

Install pre-requisite requirements.
Untar the basemap version X.Y.Z source tar.gz file, and and cd to the basemap-X.Y.Z directory.
Install the GEOS library. If you already have it on your system, just set the environment variable GEOS_DIR to point to the location of libgeos_c and geos_c.h (if libgeos_c is in /usr/local/lib and geos_c.h is in /usr/local/include, set GEOS_DIR to /usr/local). Then go to next step. If you don’t have it, you can build it from the source code included with basemap by following these steps:
```
cd geos-3.3.3
export GEOS_DIR=<where you want the libs and headers to go>
# A reasonable choice on a Unix-like system is /usr/local, or
# if you don't have permission to write there, your home directory.
./configure --prefix=$GEOS_DIR
make; make install
```
cd back to the top level basemap directory (basemap-X.Y.Z) and run the usual python setup.py install. Check your installation by running from mpl_toolkits.basemap import Basemap at the Python prompt.
To test, cd to the examples directory and run python simpletest.py. To run all the examples (except those that have extra dependencies or require an internet connection), execute python run_all.py.