Manual for ISCE installation with Anaconda / Miniconda (MacOS with intel processors)

For Apple Silicon, please follow this guide > https://github.com/lijun99/isce2-install#macosx-with-anaconda3-and-homebrew-apple-silicon

There are several ways to install ISCE on MacOS Machine with Anaconda / Miniconda.

• Install from conda-forge repositories
• Build from source code with cmake

First, a conda distribution must be installed.

There are two versions you can choose from.

1. Anaconda
   It comes with a lot of python modules preinstalled (numpy, pandas, matplotlib etc.). You can choose this if you have enough disk space. Also, it comes with anaconda-navigator which has various tools such as DataLore, JupyterLab & Notebook, Spyder IDE, VS Code, etc.
2. Miniconda
   It only installs conda environment with basic python packages. You can install all the other packages or tools later.

What should you choose? Anaconda or Miniconda?

Choose Anaconda if you:

1. Are new to conda or Python.
2. Like the convenience of having Python and over 1,500 scientific packages automatically installed at once.
3. Have the time and disk space---a few minutes and 3 GB.
4. Do not want to individually install each of the packages you want to use.
5. Wish to use a curated and vetted set of packages.

Choose Miniconda if you:

1. Do not mind installing each of the packages you want to use individually.
2. Do not have time or disk space to install over 1,500 packages at once.
3. Want fast access to Python and the conda commands and you wish to sort out the other programs later.

Installing Anaconda / Miniconda

Download either Anaconda or Miniconda:

1. Anaconda:

• You can either choose graphical or command line installer from >https://www.anaconda.com/products/individual#macos

• Direct link for 2021.05 version > https://repo.anaconda.com/archive/Anaconda3-2021.05-MacOSX-x86_64.sh

• Make sure to verify hash code after downloading. Link to check hash codes > https://docs.anaconda.com/anaconda/install/hashes/index.html

• Run shasum -a 256 filename on the terminal.

2. Miniconda:

• You can either choose graphical or command line installer from > https://docs.conda.io/en/latest/miniconda.html

• Direct link for latest version > https://repo.anaconda.com/miniconda/Miniconda3-latest-MacOSX-x86_64.sh

• Hash codes are also written on the download page.

• Use the graphical installer or run bash Miniconda3-latest-MacOSX-x86_64.sh or bash Anaconda3-2021.05-MacOSX-x86_64.sh on the terminal to install from the command line.

• Choose "yes" for prompts. This will install Anaconda / Miniconda in your home directory.

3. Checking the Installation

The installation will write conda initialize in your default shell profile. You can find the shell profile in your home directory by running:

cd  
ls -la  

Zshell > .zprofile Bash > .bash_profile

• An example for conda initialize:

# >>> conda initialize >>>  
# !! Contents within this block are managed by 'conda init' !!  
__conda_setup="$('/Users/yagizalp/anaconda3/bin/conda' 'shell.zsh' 'hook' 2> /dev/null)"  
if [ $? -eq 0 ]; then  
    eval "$__conda_setup"  
else  
    if [ -f "/Users/yagizalp/anaconda3/etc/profile.d/conda.sh" ]; then  
        . "/Users/yagizalp/anaconda3/etc/profile.d/conda.sh"  
    else  
        export PATH="/Users/yagizalp/anaconda3/bin:$PATH"  
    fi  
fi  
unset __conda_setup  
# <<< conda initialize <<<  

If you want to use conda with a different shell (bash for example) you need to run conda init.

source /<path to conda>/bin/activate  
conda init bash  

This will write conda initialize in .bash_profile.
Close and reopen the terminal and check if you have (base) at the beginning of the terminal.
(base) bash-3.2$ If the (base) is missing you may have to create .bashrc file in your home directory and cut conda initialize from .bash_profile paste it onto .bashrc. You can create .bashrc file by running:

cd  
touch .bashrc  

Run echo $PATH to check if you have anaconda / miniconda in your path.
Run which python and which python3 to make sure the python is running through anaconda / miniconda path.
If not, check your path sudo nano /etc/paths and make sure to comment other python paths.
Also check shell profiles (.bash_profile, .bashrc, .zprofile, .zshrc) and and comment python paths other than anaconda / miniconda.

4. Basic Conda Commands

When you open a new terminal you should see (base) at the beginning of the prompt. Some useful conda commands are:
conda list shows installed packages
conda info --envs shows available environments
conda create -n envName python=3.8 creates a new environment called "envName" with python 3.8
conda activate envName activates envName environment
conda deactivate envName deactivates envName environment.
conda env remove -n envName deletes envName environment.

We use environment because we can create separate python builds (different versions or with different modules) on the same machine without messing up the paths. This way you can activate which environment you want to work with and keep track of module versions, compatibility, etc.

ISCE

There are two ways to install ISCE with anaconda / miniconda.

1. Installing from conda-forge repositories.

This is very easy, but you have to add ISCE variables to your path manually.

• Create an environment called "isce2" (or you can choose whatever you like). It might be a good idea to name it the same as the isce version you are going to install.

conda create -n isce2 python=3.8

then run

conda config --add channels conda-forge
conda config --set channel_priority strict  

conda activate isce2  

conda install numpy=1.21.3

finally

conda install -c conda-forge isce2

or

conda install -c "conda-forge/label/cf202003" isce2

Conda will automatically installs all dependencies.

After the installation is finished, check if the $ISCE_HOME variable is set properly by running:

echo $ISCE_HOME
/Users/yagizalp/anaconda3/envs/isce2/lib/python3.8/site-packages/isce

The scripts that you will be using from the command line are located in the $ISCE_HOME/applications. You need to add this to your PATH if you want to call these scripts from command line.

After you activate your environment you can run:

export PATH=$ISCE_HOME/applications:$PATH

Optional

Or if you want to add this path automatically to your PATH variable once you activate your environment, go to $CONDA_PREFIX/etc/conda

There are two folders named as "activate.d" and "deactivate.d"

Go to activate.d and edit "env_vars.sh". You can add the path to $ISCE_HOME/applications in this file like this example:

export PATH=/Users/yagizalp/anaconda3/envs/isce262/lib/python3.11/site-packages/isce/applications:$PATH

However, once you deactivate your environment this path will still exist in your PATH variable unless you logout from your current shell. This may cause some path issues if you have multiple ISCE installations. In order to solve this problem you can add this line to the $CONDA_PREFIX/etc/conda/deactivate.d/env/vars.sh

export PATH=$(echo ${PATH} | sed -r 's|/Users/yagizalp/anaconda3/envs/isce262/lib/python3.11/site-packages/isce/applications:||')

2. Building ISCE with cmake.

Create an environment with python 3.8 version and activate:

conda create -n isce2 python=3.8
conda activate isce2

Install required packages:

conda install -c conda-forge git cmake cython gdal h5py libgdal pytest numpy fftw scipy basemap opencv pybind11 shapely

To compile/install mdx, you will also need:

conda install -c conda-forge openmotif openmotif-dev xorg-libx11 xorg-libxt xorg-libxmu xorg-libxft libiconv xorg-libxrender xorg-libxau xorg-libxdmcp

Optionally, if you have NVDIA graphics card, you can install CUDA compiler to speed up the process.
You will also need C/C++/Fortran compilers. Easiest way to get these to install Xcode from apple store.
Now you are ready to compile isce2.
Download source package:

mkdir -p $HOME/tools/src 
cd $HOME/tools/src 
git clone https://github.com/isce-framework/isce2.git

Compile and install isce2:

cd $HOME/tools/src/isce2
mkdir build  && cd build
cmake .. -DCMAKE_INSTALL_PREFIX=$CONDA_PREFIX -DPYTHON_MODULE_DIR=lib/python3.8/site-packages -DCMAKE_CUDA_FLAGS="-arch=sm_60" -DCMAKE_PREFIX_PATH=${CONDA_PREFIX} -DCMAKE_BUILD_TYPE=Release 
make install

Checking the Installation

Close & reopen the terminal and activate the "isce2" environment. Then run:

conda activate isce2
topsApp.py --help

This should give you an introduction text of topsApp.py module and details about its parameters.

Also

python3
import isce
isce.version.release_version

If you are not seeing any errors, installation should be fine.

For more information about ISCE, please check the github page of ISCE > https://github.com/isce-framework/isce2