Windows with ArcGIS Pro or Server

Prerequisites

64-bit x86 processor (ARM is not currently supported)
Microsoft Windows 10 or later, or Windows Server 2016 or later
ArcGIS Pro 3.2.2 or later, or ArcGIS Server 11.2 or later

Optional software

These are required to run certain parts of MGET. You can wait to install them later if desired. MGET will report detailed error messages when missing optional software is needed. Be sure to shut down all ArcGIS programs before installing them.

ArcGIS Spatial Analyst extension. This is required by a number of MGET tools produce rasters. If your ArcGIS license includes this extension but you did not install it, you can re-run the ArcGIS setup program to add it to your installation.
MATLAB Runtime R2024b (free) or the full version of MATLAB R2024b (not free). Either one is OK. These are required for front detection, larval dispersal simulation, and certain interpolation tools. You must install version R2024b; other versions will not work. Multiple versions can be installed at the same time, so if you use a different version of MATLAB for your own work, you can continue to do so, providing you install the R2024b Runtime for MGET’s use.

Windows with ArcGIS Pro installation instructions

MGET is a Python package. ArcGIS Pro utilizes conda to manage Python packages, which works best for projects that have been specifically packaged for deployment with conda. We have packaged MGET as the mget3 package on conda-forge. We recommend that ArcGIS Pro users install the conda-forge package with conda or mamba, rather than installing the corresponding mget3 package on the Python Package Index with pip.

Step 1. Install micromamba (optional but highly recommended)

You can skip this step if you have ArcGIS Pro 3.4 or later. If you have ArcGIS Pro 3.2 or 3.3, installing MGET using conda can be very slow. These versions of Pro shipped with version 4.14.0 of conda, which predates the introduction of the libmamba solver into conda. The pre-libmamba solver is notoriously slow and may spend 30 minutes or more stuck at the message “Solving environment” during MGET’s installation. It may also want to upgrade dozens of packages to the latest versions, which may be unnecessarily risky. You can work around this by first installing micromamba and using it to install MGET instead of conda. Micromamba is a stand-alone, drop-in replacement for conda. Installing it does not make any changes to your conda installation. But if you prefer to use the conda that comes with ArcGIS you can skip this step.

To install micromamba:

Start Windows PowerShell.
Open micromamba Automatic installation in your browser and copy the Windows PowerShell installation expression. It begins with Invoke-Expression.
Paste that into PowerShell and run it. If are asked “Do you want to initialize micromamba for the shell activate command?”, enter n unless you know what it means and want to do it.
Close PowerShell.

Step 2. Clone the `arcgispro-py3` environment

We strongly advise you not to install MGET or its dependencies into the default arcgispro-py3 environment that ArcGIS Pro creates when it installs. Instead:

Follow ESRI’s instructions to clone arcgispro-py3 to a new environment. In these instructions, we’ll assume your copy is called arcgispro-py3-mget. Alternatively, if you already have another environment you wish to use, you can skip this step.
Activate the new environment you created, or the existing one you want to use.

Step 3. Install MGET

Click Start, open the ArcGIS folder, and start the Python Command Prompt. It should show your desired environment as part of the command prompt, similar to this:
```
(arcgispro-py3-mget) C:\Users\Jason\AppData\Local\ESRI\conda\envs\arcgispro-py3-mget>
```
Run the following command to install the packages. Replace conda with micromamba if you installed it in Step 1:
```
conda install --channel conda-forge --yes mget3 copernicusmarine==1.3.0 aiohttp==3.9.5
```

MGET uses copernicusmarine to access Copernicus Marine Service, and it depends on aiohttp. The reason for pinning these packages to specific versions rather omitting them and allowing conda to install the latest versions is that later versions will not work with the packages that ArcGIS preinstalls and freezes into its arcgispro-py3-mget environment.

Step 4. Add the MGET toolbox to ArcGIS Pro

To use MGET’s geoprocessing tools from ArcGIS Pro, you need to add the toolbox to an ArcGIS Pro project:

Select the Insert ribbon and find the Toolbox drop-down menu. Then select Add Toolbox:

In the Add Toolbox dialog box, navigate to the folder that contains your Python environment. Typically this is a subfolder of the C:\Users\<username>\AppData\Local\ESRI\conda\envs folder. In the example below, the environment subfolder is named arcgispro-py3-mget. Inside that subfolder, navigate to Lib\site-packages\GeoEco\ArcGISToolbox. Select the file Marine Geospatial Ecology Tools.tbx and click OK:

Now you can access the toolbox from the Geoprocessing pane. After opening the Geoprocessing pane, click on Toolboxes and then drill into the toolbox to find tools of interest. Alternatively, you can search for tools by name in the Find Tools box:

Click here for some examples of using MGET’s geoprocessing tools.

Uninstalling MGET

MGET may be uninstalled like any other conda package:

Close all ArcGIS programs.
If necessary, activate the environment you want to uninstall MGET from. If that environment is already activated, you can skip this step.
Click Start, open the ArcGIS folder, and start the Python Command Prompt. It should show your desired environment as part of the command prompt, similar to this:
```
(arcgispro-py3-mget) C:\Users\Jason\AppData\Local\ESRI\conda\envs\arcgispro-py3-mget>
```
Run the following command to uninstall MGET. Replace conda with micromamba if you installed it in Step 1:
```
conda remove --yes mget3
```

Alternatively, if you no longer need the conda environment, you can just delete the environment. There is no need to uninstall MGET from it first.

Windows with ArcGIS Server installation instructions

In principle, MGET should work on ArcGIS Server so long as the prerequisite Python packages have been installed, as described above in the Windows with ArcGIS Pro installation instructions. ESRI provides some guidance on installing Python packages on ArcGIS Server for Windows in this article But we have not tested this yet so we don’t know for sure. We’ll update this documentation once we have the opportunity to try it.