2. Getting Started

UMEP is developed as a plugin for QGIS. Two different versions are available:
  • Long term release - This version is recommended for most users.
  • Development release - This version is for testing. Could be unstable.

For a more detailed description including how to install QGIS on a Windows PC (see below) or watch this instruction video. You can find more introductory videos on how to use UMEP on our YouTube-channel. UMEP has been developed using Python 2.7.x and QGIS 2.x in a Windows environment. Since QGIS is a multi-platform software system it works on other platforms as well. UMEP is still under development so there may be missing documentation and instability. Please report any issues to the code repository. Also, have a look in FAQ for further installation tips and issues.

2.1. Installation

2.1.2. Download and installation of the UMEP-plugin from within QGIS

Long-term release (Recommended)

  1. Start QGIS
  2. Go to: Plugins -> Manage and Install Plugins…
  3. Search for UMEP
  4. Click Install Plugin (or Upgrade if already have an older version installed from before).

Development release (unstable)

  1. If you have an installed version of UMEP in your QGIS, uninstall it by going to “Plugins -> Manage and Install Plugins -> Installed -> UMEP” and click Uninstall plugin
  2. To download UMEP from the repository click this link and download repository
  3. Close QGIS if open
  4. Extract the downloaded zip archive into the folder C:\Users\.qgis2\python\plugins. If the folder plugins does not exist, install any plugin using Plugins -> Manage and Install Plugins and the folder should appear.
  5. Rename the extracted folder to UMEP
  6. Start QGIS. The UMEP plugin should be visible in the QGIS toolbar. If not, go to “Plugins -> Manage and Install Plugins -> All” and search for UMEP. Make sure that you also tick in the box Show also experimental plugins in the “Settings”-tab.

Test datasets and tutorials are available to try some of the tools out.

2.2. Adding missing Python libraries and other OSGeo functionalities

Some of the plugins in the UMEP tool, for example The WATCH data plugin, requires some Python libraries such as pandas and scipy that might not been included when you installed QGIS. If so, it is necessary to install them to make this plugin work. Below are instructions on how to this for different operation systems. The same procedures can also be used to obtain other tools and functionalities from the OSGeo repository. Note: In order for scipy to work you will also need to make sure pillow is installed.

  • Operating System and Installation instructions

    1. Linux
      • Linux comes with its own Python installation which QGIS makes use of. This makes it possible to directly use pip (an installation tool of Python libraries) to add missing libraries. Simply open a terminal window and type sudo pip install pandas if you want to install this library. In order to install pip open a terminal and type: sudo easy_install pip. You might need to restart QGIS to get it to work.

      • Or refer to startup.py to modify start up file of QGIS by including the paths to pandas and scipy. An example of startup.py may look like:

        import sys
    2. Windows
      • There are two options available:
      1. As Windows has no Python installation included, QGIS make use of a separate Python installation added when QGIS was installed on your PC.

      This results in that pip cannot be used directly. However, if you installed QGIS according to the recommendations in Getting started you should have a OSGeo4W shell installed where you can use pip to add desired Python libraries. OSGeo4W shell is found in the Windows start menu. This shell is currently (October 2018) only for QGIS long term release (2.18).

      You need to run as an administrator of your PC. To do this, right-click on OSGeo4W shell and choose run as administrator. In the command window that appear, write:

      pip install pandas
      1. Installation of pandas Restart the installation (64-bit) For Advanced Users (see Getting started) and choose Advanced Install. When you come up to Select Packages search for pandas and click on Skip until you see a version number of pandas (see left picture). Finish the installation.

      This method can also be used to include other missing libraries such as gdal etc.

      pip uninstall netCDF4
      pip install netCDF4==1.2.9

      Fig. 2.1 Installation of pandas

    3. Mac OS X
      • Follow the instructions for Linux. *Note*: this approach is tested to be working under Mac OS X 10.11.5.
    4. Other Platforms
      • Other platforms require the packages to be installed to the QGIS Python path, which differs depending on operating system. Or refer to startup.py to modify start up file of QGIS by including the paths to pandas and scipy. An example of startup.py may look like

        import sys


There are currently (October 2018) some conflict issues between versions in some python libraries in the OSGeo repository (matplotlib 3.0.0 and pandas 0.23.4). To solve this there are a couple of steps you need to do in a windows environment.

First you need to create your own OSGeo4W shell but for a QGIS3 environment. This is done by creating your own .bat file. Create a text file and save as e.g. shell_QGIS3.bat including the text below. MAke sure that your OSGEO4W_ROOT is the location where you installed OSGeo.

call %OSGEO4W_ROOT%\bin\o4w_env.bat"
call qt5_env.bat
call py3_env.bat
path %OSGEO4W_ROOT%\apps\qgis\bin;%PATH%
set QGIS_PREFIX_PATH=%OSGEO4W_ROOT:\=/%/apps/qgis
set VSI_CACHE_SIZE=1000000
set QT_PLUGIN_PATH=%OSGEO4W_ROOT%\apps\qgis-dev\qtplugins;%OSGEO4W_ROOT%\apps\qt5\plugins

@echo on
@if [%1]==[] (echo run o-help for a list of available commands & cmd.exe /k) else (cmd /c "%*")

Second, run the bat file, uninstall matplotlib and reinstall with the version 2.3.3 using the command below:

pip uninstall matplotlib
pip install matplotlib==2.2.3