Run on system

to run the examples, install M²aia and set system variables to load the required libraries

follow this instructions.

1. Follow the installation instructions on pyM²aia's project site.
2. git clone https://github.com/m2aia/pym2aia-examples
3. cd pym2aia-examples
4. git submodule update --recursive --init
5. pip install -r requirements.txt

By running an example, the required data set MTBLS2639 will be downloaded automatically. This will take some time (e.g. in Example I only one slice of four is downloaded, that takes about 12 minutes).

Run in Docker

Prepare source files

1. git clone https://github.com/m2aia/pym2aia-examples
2. cd pym2aia-examples

Build without gpu support (Example I-III)

1. docker build -t pym2aia-examples -f Dockerfile .
2. docker run -ti --rm -v $(pwd):/examples pym2aia-examples Example_I_ImzMLMetaData.ipynb $(id -u $USER)

Build with gpu support (Example IV-VI)

1. docker build -t pym2aia-examples -f Dockerfile.gpu .
2. docker run -ti --rm --gpus all -v $(pwd):/examples pym2-gpu Example_IV_A_AutoEncoder_IndividualModels.ipynb $(id -u $USER)

if the last argument $(id -u $USER) is set, all items in $(pwd)/data, $(pwd)/results and $(pwd)/models will change ownership to the current user (otherwise files will be owned by the root user).