STRAWb

STRAWb is a python module which simplifies and streamlines the work with the STRAWb data. Beside this Readme, there is more documentation at the repository Wiki page.

Installation

For the installation you have two options.

Installation directly from the repository

Pip can directly install the package from the repository. As pip only compares the version number and not the code, uninstall an existing installation before you install it from the repository. For updating the package just rerun the same commands. You can also specify the branch by changing master accordingly. The commands are:

pip3 install -U git+https://github.com/pone-software/strawb_package.git@master  # Install it from the repository

Afterwards, copy the config file to your home directory at ~/strawb, e.g. with:

makdir ~/strawb
vim ~/strawb/config

c/p, adopt the paths to your setup if needed and save (:wq).

Updating with pip

Pip will compare the version of the package and only update if the version number has changed (defined in setup.py). Only a few commits will increase the version. To get the update with the lates commits, uninstall and install strawb again with:

pip3 install --no-deps --ignore-installed git+https://github.com/pone-software/strawb_package.git@master

or

pip3 uninstall -y strawb  # Uninstall an existing installation
pip3 install -U git+https://github.com/pone-software/strawb_package.git@master  # Install it from the repository

Installation for developers

This installation downloads the source code, and the package loads directly from the source code for every import. Therefore, any changes to the code will have direct effect after an import.

Go to the directory of this README you are reading is placed (basically, to the directory of the pyproject.toml file, but this should be the same). Depending on your Python installation adopt python3/pip3 to python/pip, however python3 is required. Run:

mkdir /path/to/repros  # adopted the path, be aware that git clone creates a directory with the repository name
cd /path/to/repros # enter the directory

# clone/download the repository
git clone https://github.com/pone-software/strawb_package.git  # downloads the repository
cd strawb_package  # enter the repository directory

# install the package
python3 -m build  # This will create the files located in the folder `.egg-info`
pip3 install -U -r requirements.txt  # install the required python packages
pip3 install -U --user -e .  # install the package in developer mode.

Update repository when output of notebooks prevent a git pull

This repository holds several notebooks. Git commit is configured to clear all notebooks outputs before committing to include code only and significantly reduce the size of the notebooks. When you run a notebook, the notebook will change also if you don't modify any cell due to the output. The msg will be:

git pull
# ...
# Please commit your changes or stash them before you merge.
# Aborting

To do a git pull, you have to git reset --hard origin/master or the branch you want to choose instead of master. Be aware, this will reset your local repository to the lattes commit and your changes to any file in the repository will be lost. If you want to keep the changes, follow the usual commit-merge-path.

Afterwards a git pull will work fine again. To prevent this behaviour, you can make local copies of the notebooks and run them there. Once you want to upload changes of a notebook, copy the notebook back and do a git commit (+ git push).

Known issued at installation

• Missing hdf5 installation If you see an error like

...
  Loading library to get build settings and version: libhdf5.so
  error: Unable to load dependency HDF5, make sure HDF5 is installed properly
  error: libhdf5.so: cannot open shared object file: No such file or directory
  ----------------------------------------
  ERROR: Failed building wheel for h5py
 ...

hdf5 isn't installed. On Linux/Ubuntu, run

sudo apt-get install libhdf5-dev

• Anaconda installation

If you use a Python-Anaconda installation, you have to install opencv-python manually as Anaconda hasn't listed opencv-python and therefore can install it. Activate your environment in a terminal (conda activate) and run (or with sudo if this doesn't work):

pip install opencv-python

• AssertionError: Egg-link

In case you see an AssertionError: Egg-link, run:

rm ~/.local/lib/pythonX.X/site-packages/H5DAQ.egg-link

where X.X is your python version, e.g., 3.7 and run the following command again.

pip3 install -U -e .

This problem happens if you move the source code directory after you created the egg-link file.

TODO List:

• add readme for every sensor typ
• add examples
• add all sensor types