MCELite is a multi-criteria evaluation (MCE) tool written in Python. Given the appropriate raster layers as input, it executes one of three types of MCE: Boolean overlay, weighted linear combination (WLC) or ordered weighted averaging (OWA). The installation package includes documentation and examples.

MCELite is designed to work in four modes: as a standalone Python library that can be called from other Python programs, as an ArcMap 9.3 plugin, as an IDRISI Taiga plugin and as a QuantumGIS (QGIS) plugin. A command line script for running MCELite ( is included with the library. Each specific installation includes an “mcelite” package in a sub-directory.

Currently, only the QGIS plugin is available while bugs are worked out, and can be downloaded through the QGIS plugin manager interface (it is marked as experimental) from the QGIS Contributed respository, or from github. The ArcGIS plugin will be made available later (if you want to try it, just email me).


MCELite requires, in addition to the standard libraries included with Python, that GDAL with Python bindings be installed, as well as the Python numeric extension module, NumPy. GDAL is included with all three of the software packages for which MCELite functions as a plugin, and the NumPy module is installed with QGIS and ArcMap. However, GDAL is not available to Python in either IDRISI or ArcMap, and needs to be installed separately. GDAL, of course, must be installed independently to take advantage of the standalone functionality of MCELite.

Analytical Functions

There are several analytical functions that can be executed by MCELite. The package was designed as a work-alike to the IDRISI MCE tool (with the addition of the sensitivity analysis feature). However, MCELite does not have functionality for developing fuzzy membership criteria or factor weights. Factor files will have to be developed using other tools and the raster layers created used as input to the MCELite tool. Currently MCELite provides three types of analysis: boolean overlay, weighted linear combination (WLC) and ordered weighted averaging (OWA).