Skip to content

Latest commit

 

History

History
270 lines (190 loc) · 11.3 KB

installation.md

File metadata and controls

270 lines (190 loc) · 11.3 KB
Raw

1. Install the released version

via conda / mamba

MintPy is available on the conda-forge channel. The latest released version can be installed via conda as:

conda install -c conda-forge mintpy

or via mamba as:

mamba install -c conda-forge mintpy

or via docker

Docker allows one to run MintPy in a dedicated container, which is essentially an efficient virtual machine, and to be independent of the platform OS. First, install docker if you have not already done so. Then run the following to pull the latest stable released container image version from MintPy GitHub Container Registry to your local machine:

docker pull ghcr.io/insarlab/mintpy:latest

Check docker.md for more details on Docker container image usage, e.g. pulling the development version and running in a shell or Jupyter server.

or via apt (for Linux Debian)

MintPy is available in the main archive of the Debian GNU/Linux OS. It can be installed by using your favorite package manager or running the following command:

apt install mintpy

The same procedure, in principle, can be used in Ubuntu and all the other Debian derivatives. Check the Debian GIS Project page for more detailed usage.

2. Install the development version

Note: The installation note below is tested on Linux and macOS, and is still experimental on Windows (may have bugs).

MintPy is written in Python 3 and relies on several Python modules, check the requirements.txt file for details. We recommend using conda or macports to install the python environment and the prerequisite packages, because of the convenient management and default performance setting with numpy/scipy and pyresample.

2.1 Install on Linux

Click to expand for more details

a. Download source code

Run the following in your terminal to download the development version of MintPy:

cd ~/tools
git clone https://github.com/insarlab/MintPy.git

b. Install dependencies via conda

Install miniconda if you have not already done so. You may need to close and restart the shell for changes to take effect.

# use wget or curl to download in the command line or click from the web browser
# for macOS, use Miniconda3-latest-MacOSX-x86_64.sh instead.
wget https://repo.anaconda.com/miniconda/Miniconda3-latest-Linux-x86_64.sh
bash Miniconda3-latest-Linux-x86_64.sh -b -p ~/tools/miniconda3
~/tools/miniconda3/bin/conda init bash

Install the dependencies into a custom existing environment [recommended] by running:

# To speedup, try "conda install mamba", then use "mamba install" to replace "conda install" below

# Add "gdal'>=3'" below to install extra dependencies if you use ARIA, FRInGE, HyP3 or GMTSAR
# Add "isce2"     below to install extra dependencies if you use ISCE-2
conda install -c conda-forge --file ~/tools/MintPy/requirements.txt

Or install the dependencies to a new environment, e.g. named "insar", by running:

# Add "gdal'>=3'" below to install extra dependencies if you use ARIA, FRInGE, HyP3 or GMTSAR
# Add "isce2"     below to install extra dependencies if you use ISCE-2
conda create --name insar --file ~/tools/MintPy/requirements.txt
conda activate insar

c. Install MintPy

via pip [recommended]

We recommend installing MintPy in the "editable" mode. This mode installs the package without copying files to your interpreter directory (e.g. the site-packages directory), thus, one could "edit" the source code and have changes take effect immediately without having to rebuild and reinstall.

python -m pip install -e ~/tools/MintPy

or via path setup

Add below in your source file, e.g. ~/.bash_profile for bash users or ~/.cshrc for csh/tcsh users:

if [ -z ${PYTHONPATH+x} ]; then export PYTHONPATH=""; fi
export MINTPY_HOME=~/tools/MintPy
export PATH=${PATH}:${MINTPY_HOME}/src/mintpy/cli
export PYTHONPATH=${PYTHONPATH}:${MINTPY_HOME}/src

2.2 Install on macOS

Click to expand for more details

Install Xcode with command line tools, if you have not already done so.

  • Install Xcode from App store

  • Install command line tools within XCode and agree to the terms of license.

    • xcode-select --install -s /Applications/Xcode.app/Contents/Developer/
sudo xcodebuild -license
  • Install XQuartz, then restart the terminal.

Install MintPy via conda, which is the same as the instruction for Linux.

Or install MintPy via MacPorts

Same as the instruction for Linux, except for the "b. Install dependencies" section, which is as below. Note that the installation procedure via MacPorts has not been maintained since Sep 2022, and will likely be phased out at some point, since conda/mamba works seamlessly on both MacOS and Linux.

Install macports if you have not done so. Add the following at the bottom of your ~/.bash_profile file:

# MacPorts Installer addition on 2017-09-02_at_01:27:12: adding an appropriate PATH variable for use with MacPorts.
export PATH=/opt/local/bin:/opt/local/sbin:${PATH}
export MANPATH=/opt/local/share/man:${MANPATH}
# Finished adapting your PATH environment variable for use with MacPorts.

Update the port tree with the following command. If your network prevents the use of rsync or svn via HTTP of the port tree, try Portfile Sync via a Snapshot Tarball.

sudo port selfupdate

Install the dependencies by running the following:

# install dependencies with macports
# use "port -N install" to use the safe default for prompt questions
sudo port install $(cat MintPy/docs/ports.txt)

# install dependencies not available on macports: pysolid, pykml, pykdtree, pyresample, cdsapi
sudo -H /opt/local/bin/pip install git+https://github.com/insarlab/PySolid.git
sudo -H /opt/local/bin/pip install git+https://github.com/insarlab/PyAPS.git
sudo -H /opt/local/bin/pip install git+https://github.com/tylere/pykml.git
sudo -H /opt/local/bin/pip install git+https://github.com/storpipfugl/pykdtree.git
sudo -H /opt/local/bin/pip install git+https://github.com/pytroll/pyresample.git
sudo -H /opt/local/bin/pip install git+https://github.com/ecmwf/cdsapi.git

2.3 Install on Windows

Click to expand for more details

Same as the instruction for Linux, except for the "c. Install MintPy" section, only the pip install approaches are recommended, as the path setup approach is not tested.

3. Post-Installation Setup

a. ERA5 for tropospheric correction

Set up an account for ERA5 to download weather re-analysis datasets for tropospheric delay correction as described in insarlab/PyAPS.

WEATHER_DIR: Optionally, if you defined an environment variable named WEATHER_DIR to contain the path to a directory, MintPy will download the GAM files into the indicated directory. Also, MintPy will look for the GAM files in the directory before downloading a new one to prevent downloading multiple copies if you work with different datasets that cover the same date/time.

b. Dask for parallel processing

We recommend setting the temporary-directory in your Dask configuration file, e.g. ~/.config/dask/dask.yaml, by adding the following line, to avoid the potential workspace lock issue. Check the dask.md file for more details on parallel processing.

temporary-directory: /tmp  # Directory for local disk like /tmp, /scratch, or /local

# If you share the same machine with others, use the following instead to avoid permission issues.
# temporary-directory: /tmp/{replace_this_with_your_user_name}

c. Extra environment variables setup

We recommend setting the following environment variables, e.g. in your ~/.bash_profile file, to avoid occasional errors with GDAL VRT and HDF5 files I/O.

export VRT_SHARED_SOURCE=0             # do not share dataset while using GDAL VRT in a multi-threading environment
export HDF5_DISABLE_VERSION_CHECK=2    # suppress the HDF5 version warning message (0 for abort; 1/2 for printout/suppress warning message)
export HDF5_USE_FILE_LOCKING=FALSE     # request that HDF5 file locks should NOT be used

d. Tab complete sub-commands

We recommend activating the tab completion as below if you:

  1. Use bash or zsh shells
  2. On Linux or macOS
  3. Run MintPy in sub-commands style, e.g. mintpy view instead of view.py
activate-global-python-argcomplete --user

On macOS, there is a bash version compatibility issue, which can be fixed as below. If you use the default zsh shell, no fix is needed.

Click to expand for more details

macOS ships with an older version of the bash shell (run echo $BASH_VERSION to check the version number), but argcomplete requires the newer version (>=4.2), which can be installed using Homebrew as:

brew install bash

You will also need to add the bash to /etc/shells and change your shell via chsh as:

echo "/usr/local/bin/bash" | sudo tee -a /etc/shells
chsh -s "/usr/local/bin/bash"

Then run activate-global-python-argcomplete --user again. You may need to add the following to your ~/.bash_profile file.

# bash completion
if [ -f ~/.bash_completion ]; then
    . ~/.bash_completion
fi