Getting started¶

Installing¶

Installing from PyPI (platform-independent)¶

Execute pip install pymepix. This should install pymepix including all dependencies.

Installing from git source directly (platform-independent)¶

You can clone pymepix from our main git repository:

git clone https://gitlab.desy.de/CMI/CMI-public/pymepix.git

Navigate into the pymepix library (cd pymepix) and run python -m pip install .

Build Documentation¶

To build the documentation for pymepix locally perform the following commands. The first line is only required if there are changes in the package structure or new classes or packages have been added. To only build the existing documentation only the second line must be executed.

sphinx-apidoc -o ./doc/source/ ./pymepix
sphinx-build -b html doc/ public/

Adapt pymepix/config/default.yaml according to your setup.

timepix:
ip: ‘127.0.0.1’ # TCP address of the timepix camera
port: 50000 # Port for TCP communication with timepix camera
udp_ip1: ‘127.0.0.1’ # IP address on which data is received from timepix 3 and 4
udp_ip2: ‘127.0.0.11’ # IP address on which data is received from timepix 4
data_port: 8110 # UDP Port receives data on this port
sophy_config: ‘path_to_sophy_config’ # ASI SPIDR SoPhy config
camera_generation: 3 # 3 or 4
zmq_port: 25000 # Port for pipeline ipc with ZeroMQ
trainID: (will soon be deprecated)
connected: False
device: ‘/dev/ttyUSB0’

Dependencies¶

The majority of pymepix only depends on numpy. To use centroiding, the scikit-learn package is required

numpy
scikit-learn: Centroiding and data reduction (Using DBSCAN algorithm for clustering)
scipy: Calculation of the centroids properties from the identified clusters
pyzmq: Inter process communication in the processing pipeline
h5py: Saving processed data as hdf5 files
tqdm: Display a progessbar for post processing
pyyaml: Konfiguration of camera (ip, port, …)
pyserial (optional): Only used for inclusion of USBTrainID at FLASH and XFEL