MTPy installation guide for Windows 10 and Ubuntu 18.04
MTPy has a PyPi release. In the installation guide below, it's possible to substitute installing from the repository with installing from PyPi using:
pip install mtpy
At the current time this is not recommended as the PyPi release is outdated. It is recommended to install from the repository to get the latest development code.
This guide will explain how to install MTPy on a Windows 10 system using Anaconda and Python 3.7.
- Install Anaconda Python 3.7 from https://www.anaconda.com/distribution/
If you are comfortable using a terminal, the installation can be completed using an Anaconda Prompt.
- Open Anaconda Prompt
- Create a new environment using
conda create --name mtpy - Activate the environment using
conda activate mtpy - Install required packages:
conda install gdal libgdal geopandas netcdf4 pyyamlpip install obspy
- Install mtpy using the instructions in the GUI section
Alternatively, the installation of required packages can be completed using the Anaconda GUI.
- After installing Anaconda, open the Anaconda Navigator app
- In the left sidebar, select Environments, then at the bottom of the window select Create
- Give your new environment a suitable name and select Python 3.7 as the package, then press the green Create button to confirm:
- Select the environment you have created from the list of available environments
- In the packages window to the right, select Not installed from the drop down and enter gdal in the search field
- Select the checkboxes next to gdal and libgdal, then click the Apply button in the lower right corner
- A window will display confirming dependencies to install, click Apply
- Repeat the process for the following additional packages:
- geopandas
- netcdf4
- pyyaml
- For the next step, click the arrow next to your environment name select Open Terminal
- You are now in a command prompt with your created Python environment active
- Run the command
pip install obspy
This can be completed in the Anaconda terminal, or your preferred Windows git client (e.g. git bash)
- Change directory to the location where you want the mtpy code to be cloned to
- The Anaconda terminal is a Windows command prompt, so type
DRIVE_LETTER:to change to the appropriate drive andcd path\to\directoryto change directories.
- The Anaconda terminal is a Windows command prompt, so type
- Run the command
git clone https://www.github.com/MTgeophysics/mtpy - Once cloned, move into the cloned mtpy directory with
cd mtpy
Now that mtpy has been installed into your environment, you can launch applications from Anaconda Navigator from within this environment and they will have access to mtpy.
- Click Home on the left sidebar of Anaconda Navigator
- In the Applications on drop down menu, select your created environment
- Install and launch your desired software
For example, you can run mtpy in Spyder through Anaconda Navigator:
Note: if you have previously used mtpy with Windows, you may be used to putting os.chdir('/path/to/mtpy') in your scripts. Now that mtpy is installed as a package in your environment, you no longer need to do this. So long as you launch your application using the created environment, you can import mtpy like any other package without changing directory.
- Open an Anaconda Prompt in your mtpy environment (or your preferred git client, but ensure your mtpy environment is active).
- Change directory to the mtpy directory containing the mtpy code
- Checkout the desired branch with
git pull origin <branch-name>(e.g.git checkout develop) - Run the command
git pull origin <branch-name>- If you have local changes that are preventing you from switching branches or pulling new code, and you wish to keep them, first run
git stash. This will stash your changes and clean your working directory. Once you have changed branches and updated the code, you can recover your changes usinggit stash pop, which will reapply them. - If you don't mind losing the changes, run
git checkout -- .to revert your changes and then checkout and pull the new code.
- If you have local changes that are preventing you from switching branches or pulling new code, and you wish to keep them, first run
- Reinstall mtpy by running the command
pip install .- Tip: if you are frequently pulling code changes, or making your own changes to mtpy, then install with
pip install -e .. The-etells pip to install in editable mode, so changes will automatically be incorporated into the package without having to reinstall after every change.
- Tip: if you are frequently pulling code changes, or making your own changes to mtpy, then install with
You can now run MTPy workflows as described in the user guide.
These instructions have been tested with Ubuntu 18.04. If you are using a different distribution, the name of your package manager may be different as well as available package versions.
mtpy can be installed on Linux with or without Anaconda. If you're unsure, it's best to use Anaconda as it will handle OS dependencies for you.
If you use a GUI desktop environment with your Linux OS, then you can download Anaconda for your OS (https://www.anaconda.com/distribution/) and follow the Anaconda GUI instructions above.
If you are running Ubuntu server or prefer to use the terminal, follow these instructions:
- Visit https://repo.anaconda.com/archive/ to find the latest or desired Linux install script (e.g. Anaconda3-2020.02-Linux-x86_64.sh)
- Download the script with
curl -O https://repo.anaconda.com/archive/Anaconda3-2020.02-Linux-x86_64.sh(substituting the script for your desired version) - Run the script with
./Anaconda3-2020.02-Linux-x86_64.shand follow the prompts - Create a new environment using
conda create --name mtpy - Activate the environment using
conda activate mtpy - Install required packages:
conda install gdal libgdal geopandas netcdf4 pyyamlpip install obspy
- Change directory to the location where you want the mtpy code to be cloned to
- Run the command
git clone https://www.github.com/MTgeophysics/mtpy - Once cloned, move into the cloned mtpy directory with
cd mtpy - Run the command
pip install .- This command installs the mtpy code as a package in your created Python environment
If you prefer to manage your own Python environments and packages, mtpy can be installed without Anaconda.
You can choose to install Python 3.7 from source, or follow these instructions to install from the deadsnakes ppa:
sudo apt updatesudo apt install software-properties-common-
sudo add-apt-repository ppa:deadsnakes/ppa- This command adds the deadsnakes repository to your package manager. This repository maintains versions of Python for Ubuntu that aren't available by default.
sudo apt install python3.7
sudo apt install gdal-binsudo apt install libgdal-dev-
sudo apt install build-essential- This will install gcc, required for building some of mtpy's Python dependencies
- Add GDAL to your
CPLUS_INCLUDE_PATHandC_INCLUDE_PATHexport CPLUS_INCLUDE_PATH=/usr/include/gdal:$CPLUS_INCLUDE_PATH-
export C_INCLUDE_PATH=/usr/include/gdal:$C_INCLUDE_PATH- Add these commands to your
.bashrcso they are automatically applied when you start a new shell
- Add these commands to your
- Change directory to where you want your mtpy Python virtual environment to exist
- Create the environment using
python3.7 -m venv mtpy - Activate the environment using
source mtpy/bin/activate - Upgrade pip and setuptools with
pip install -U pip setuptools - Install dependencies:
pip install geopandas netcdf4 obspy pyyaml gdal
Follow the Using Anaconda in the terminal mtpy install instructions above.
Follow the Updating mtpy instructions in the Windows section above. You can use an Anaconda Prompt if you are running Anaconda Navigator, or substitute its use with your Linux terminal (ensure your mtpy environment is active!).
You can now run MTPy workflows as described in the user guide.