Parallelizing PIE analysis¶

Although PIE does not automatically support multithreading, it is set up to easily be run via batch submission. To do this, the colony tracking step for each individual xy position is separated from the steps that compile the data across positions and calculate growth rates.

Analyzing individual imaged xy positions¶

To analyze each imaged position individually (e.g. via batch jobs), you need to pass the integer corresponding to the xy position being analyzed and the path to the setup file. This can be done e.g. for position 11 of the SL_170619_2_GR_small sample data:

pie track_single_pos 11 /local/path/to/PIE/sample_PIE_setup_files/gr_phase_setup.csv

import PIE
PIE.track_colonies_single_pos(
    11,
    '/local/path/to/PIE/sample_PIE_setup_files/gr_phase_setup.csv'
    )

Combining position-wise data and calculating growth rates¶

After every imaged xy position has been analyzed, the data can be combined and growth rates can be calculated by simply running PIE.run_timelapse_analysis(...) (or pie run_timelapse_analysis ...), as in Running PIE timelapse experiments, with the default repeat_image_analysis_and_tracking=False option.