Installation¶
Requirements¶
Python Packages:
Python 3.7 (or later)
pandas
numpy (1.18)
matplotlib
seaborn
pytorch
scikit-learn
xlsxwriter
wxpython (4.0.7)
networkx
pypubsub
python.app (for MacOS X)
setuptools_scm
Creating a Conda environment¶
The code Git repository https://github.com/uvaKCCI/flimanalyzer.git contains an environment.yml file defining required Python packages. The python.app package, only required for MacOS X, is not included in the environment file and has to be installed manually, see below.
Run the following commands to create a flimenv Conda environment based on the environemnt.yml file:
git clone https://github.com/uvaKCCI/flimanalyzer.git
cd flimanalyzer
conda env create -f environment.yml
conda activate flimenv-0.4.0
python setup.py install
Installations on Windows and MacOS require a couple additional steps.
Windows Installation
Install VS Code.
MacOS X Installation
Activate the Conda environment, install the MacOS specific python.app package, and patch the shebang of the flimanalyzer console script.
conda activate flimenv-0.4.0
conda install python.app
sed -i '' -e 's/bin\/python/python\.app\/Contents\/MacOS\/python/' $(which flimanalyzer)
Run the application¶
From the command line
On Mac and Linux, run this command:
conda activate flimenv-0.4.0
export PREFECT__FLOWS__CHECKPOINTING=true
flimanalyzer
On Windows, run this command:
conda activate flimenv-0.4.0
set PREFECT__FLOWS__CHECKPOINTING=true
flimanalyzer.exe
On Powershell, run this command:
conda activate flimenv-0.4.0
$env:PREFECT__FLOWS__CHECKPOINTING='true'
flimanalyzer.exe
From Anaconda Spyder
Launch the Anaconda Navigator
In the
Applications ondrop-down, select the name of the conda environment, e.g.flimenv.Launch Spyder (you may have to install it in your custom flimenv environment).
In the Spyder editor, open the
analyzerapp.pyfile.Go to
Run–>Configuration per file...and check the boxExecute in external system terminal.Go to
Run–>Runor click the button with the green triangle in the Spyder toolbar.