Extract Physical Values

Generate and store the best fit values and metadata for an experiment.

Quick Start

Once the Fourier terms have been calculated by process_images.py this script fits the data to the experimental spectrum and creates a fitting_vals.csv table for the experiment directory. After this use create_plots.py to merge the results from several experiments and create summaries of the data.

Outline

For each Fourier terms file we fit every granule to a given theoretical spectrum. From this we calculate the best estimates for the physical values, typically the surface tension and bending rigidity.

This is saved in a summary table, with other metadata about the granules, this includes properties of the granule such as its size or the time since treatment, but also abstract parameters, such as the quality of fitting to the expected spectrum.

Once the summary table is created, we add a number of secondary columns for convenience, typically physical values re-scaled by some factor or on a log scale. We also bin the time values into 5 minute intervals that can be used for aggregation/averaging.

As the fitting is somewhat expensive, we provide an option for reusing the summary table . This can be cleared by using the --clear flag, and new data will be generated.

Merge Caches

Once we have gathered all of the data in an experiment directory, we can then analyse these files together using the merge_caches function. We provide means for dealing with mismatched or missing columns.

Simply provide a list of Path objects pointing to the fittingVals.csv files to merge.

Flags

--all-frames

Do not remove frames that have failed the boundary check when fitting for a granule. The pass rate is still recorded with the granule and may be used to filter the entire granule from later analysis.

-j, --processes1

Number of cores to run the fittings on; with each core running a different time series. Defaults to single core for debugging, but multiple cores are recommended.

--clear

Recreate the fittingVals.csv file.

flickerprint.workflow.extract_physical_values.main(working_dir: Path, plotting=False, cores=1)

Merge multiple Fourier terms into a single file.

flickerprint.workflow.extract_physical_values.process_fourier_file(input_path: Path, output: Path, plotting=True)

Perform spectrum fitting on one h5 file.