Skip to content

UseMacOSCondaBinary.md

Latest commit

 

History

History
142 lines (100 loc) · 5.25 KB

UseMacOSCondaBinary.md

File metadata and controls

142 lines (100 loc) · 5.25 KB

Installing, testing and using a CloudComPy binary on MacOS, with conda

Note: CloudComPy versions released in 2024 and earlier were built using a Conda Python 3.10 environment: CloudComPy310. Now, new versions are built with a Python 3.11 Conda environment: CloudComPy311

The binary CloudComPy_Conda311_MacOS-date-.zip available here is built with a Conda environment.

This binary works only on macOS Apple arm64 architecture (not on Intel processors), on recent macOS versions, not anywhere else!

Built and tested on macOS Sonoma 14.7. Please post issues on CloudComPy GitHub or GitLab in case of problem

The macOS binary provides CloudCompare and CloudCompy (same as binaries for Windows and Linux).

As CloudComPy is under development, these instructions and the link are subject to change from time to time...

CloudCompare works as it is (no specific environment). It is located in CloudComPy311/CloudCompare/CloudCompare.app and can be launched from the Finder.

CloudComPy needs a Python 3.11 configuration with at least the following packages, either with conda or not:

numpy
scipy
requests
psutils
matplotlib

You can create an environment for CloudComPy with conda, from the terminal (here, I chose to activate conda environment on demand: please adapt the instructions to your installation) The following package list corresponds to the building environment, but you can adjust the list (at least the above list):

conda activate
conda update -y -n base -c defaults conda

If your environment CloudComPy311 does not exist or to recreate it from scratch: (note: it's best to recreate the environment from time to time, as conda packages are often updated)

conda create --name CloudComPy311 python=3.11
   # --- erase previous env with the same name if existing

Add or update the packages:

conda activate CloudComPy311
conda config --add channels conda-forge
conda config --set channel_priority flexible
conda install -y "boost=1.84" "cgal=5.6" cmake "draco=1.5" "ffmpeg=6.1" "gdal=3.8" jupyterlab laszip "matplotlib=3.9" "mysql=8" notebook numpy "opencv=4.9" "openssl>=3.1" "pcl=1.14" "pdal=2.6" "psutil=6.0" pybind11 quaternion "qhull=2020.2" "qt=5.15.8" scipy sphinx_rtd_theme spyder tbb tbb-devel "xerces-c=3.2" xorg-libx11

Unzip the binary tarfile in the directory of your choice.

Using CloudCompare and CloudComPy:

CloudCompare is located in CloudComPy311/CloudCompare/CloudCompare.app and can be launched from the Finder.

CloudcomPy requires to set the Python environment and the PYTHONPATH.

If you have configured the conda CloudComPy311 environment, use the script bin/condaCloud.zsh. It overrides the conda command for activation and deactivation, it must be sourced.

From a new prompt (replace <path install> by its value):

. <path install>/bin/condaCloud.zsh activate CloudComPy311

if conda is unknown, execute the following instruction before:

. <conda_dir>/etc/profile.d/conda.sh

where <conda_dir> is the installation directory of conda (often ~/anaconda3 or ~/miniconda3)

Have a look on the script usage:

. <path install>/bin/condaCloud.zsh

If you use another Python configuration (without conda), you need to set the PYTHONPATH:

export PYTHONPATH=${CLOUDCOMPY_ROOT}:${PYTHONPATH}
export PYTHONPATH=${CLOUDCOMPY_ROOT}/CloudCompare/CloudCompare.app/Contents/Frameworks:${PYTHONPATH}
export PYTHONPATH=${CLOUDCOMPY_ROOT}/doc/PythonAPI_test:${PYTHONPATH}

where CLOUDCOMPY_ROOT is the path of CloudComPy311.

To execute a Python script (for instance myscript.py) using CloudComPy:

python myscript.py

The IDE Spyder and Jupyter can be launched in this environment:

spyder

The first time you run Spyder, you may need to add the paths to CloudCompare and the test scripts to the PYTHONPATH, using the menu Tools / PYTHONPATH Manager. These paths are the same as those defined in the condaCloud.zsh script.

jupyter notebook

An example of notebook is provided in doc/samples/histogramOnDistanceComputation.ipynb.

Execute all the Python tests:

. <path install>/bin/condaCloud.zsh activate CloudComPy311
cd  <path install>/doc/PythonAPI_test

To execute all the tests (takes about five minutes, creates about 2GB of data files):

ctest

The files created with the tests are in your user space: ${HOME}/CloudComPy/Data

Current status and problems:

  • The plugins QRDB, QSTEP, QTREEISO are not available.
  • test042.py fails on macOS

In case of problem:

There may be differences in the versions of conda packages. When updating the conda configuration, the package versions may change slightly. This is usually not a problem, but since the CloudComPy binary is fixed, there may be a version difference on a package, which makes it incompatible with CloudComPy. For your information, here is the list of package versions when CloudComPy was built.

The result of conda list command is provided in the doc directory of the binary package or, in the sources, in building directory