sensospot_parser/README.md

Sensospot Data Parser

Parsing the numerical output from Sensovation Sensospot image analysis and some other useful functions for working with the data.

Example:

    import sensospot_data

    # read the raw data of a folder
    raw_data = sensospot_data.parse_folder(<path to results directory>)

    # apply an exposure map to add more data:
    #   key relates to column "Exposure.Id"
    #   values are (Exposure.Channel, Exposure.Time)
    exposure_map = {
        1: ("Cy3", 100),
        2: ("Cy5", 150),
        3: ("Cy5", 15),
    }
    enhanced_data = sensospot_data.apply_exposure_map(raw_data, exposure_map)

    # split the measurement according to channels
    channels = sensospot_data.split(enhanced_data "Exposure.Channel")

    # merge the two cy5 measurements together, creating an extended dynamic range
    cy5_xdr = sensospot_data.create_xdr(channels["cy5"], normalized_time=25)

Avaliable functions:

from .parser import parse_file, parse_folder # noqa: F401

• parse_folder(path_to_folder) Searches the folder for parsable .csv files, parses them into one big pandas data frame and will add additional meta data from parameters folder, if it is present.
• parse_file(path_to_csv_file) Parses the csv file into a pandas data frame and will add additional some meta data from the file name. Is internally also used by parse_folder()
• split(data_frame, column)
  Splits a data frame based on the unique values of a column. Will return a dict, with the unique values as keys and the corresponding data frame as value
• apply_map(data_frame, map, idex_col) Adds information provided in the nested dictionary map to a data frame, based on the values in the data_frame column index_col.
• apply_exposure_map(data_frame, exposure_map) Adds information about the channel and exposure time to a data frame, based on the exposure id. Will get bonus karma points, if the named tuple ExposureInfo is used: {1:ExposureInfo("Cy3", 100), 2:ExposureInfo("Cy3", 100), }
• ExposureInfo(exposure_channel, exposure_time)
  A named tuple for defining an exposure map. Usage will increase readability and karma points.
• blend(data_frame, [column="Spot.Saturation", limit=2])
  If provided with a data frame with multiple exposure times for the same exposure channel, the function will blend theese two times together based on given column and limit.
• normalize_values(data_frame, [normalized_time=None])
  Adds new columns to the data frame with intensity values recalculated to the normalized exposure time. If no time is given, the max exposure time is used.
• create_xdr(data_frame, [normalized_time=None, column="Spot.Saturation", limit=2])
  This combines the methods blend() and normalize_values() into one call. What a joy!

CLI

For the (propably) most important function, there is even a cli command

Usage: parse_sensospot_data [OPTIONS] SOURCE

Arguments:
  SOURCE:             Folder with Sensospot measurement

Options:
  -o, --outfile TEXT  Output file name, relative to SOURCE, defaults to 'raw_data.h5'
  --help              Show this message and exit.

Development

To install the development version of Sensovation Data Parser:

git clone https://git.cpi.imtek.uni-freiburg.de/holgi/sensospot_data.git

# create a virtual environment and install all required dev dependencies
cd sensospot_data
make devenv

To run the tests, use make tests (failing on first error) or make coverage for a complete report.