checkModels: checkModels

In AndersenLab/easyXpress: A Package to Read, Process, and Analyze Worm Data Generated from CellProfiler’s WormToolbox

checkModels

Description

This function will create an one or more arrays of image overlays to help users evaluate cellprofiler model performance and object length thresholds.

Usage

checkModels(
  data,
  ...,
  modelName = "MDHD",
  OF = "filter",
  length_thresh = 164.685,
  strainN = 4,
  wellN = 4,
  proc.img.dir,
  well.label,
  out.dir
)

Arguments

data	A data frame output from the modelSelection function or any OF function from easyXpress.
...	<['dynamic-dots'][rlang::dyn-dots]> Variable(s) used to group data. Variable names in data are supplied separated by commas and without quotes. For example drug, concentration_um.
modelName	The name of the cellprofiler model to check in the diagnostic plots. Typically this is the smallest model in the dataset. The default is set to "MDHD".
OF	Select one of "filter", "ignore". The default is "filter", which will filter out all objects flagged by OFs. "ignore" will include all objects in the diagnostic plots.
length_thresh	An object length threshold in um used to label objects from the data. The default setting is 164.685 um. This is the standard threshold used for the AndersenLab 2X objective images. Please adjust if necessary.
strainN	The maximum number of strains to include in a overlay. This function will return at maximum strainN strains per overlay array. These strains are chosen because they have the lowest average object length in the group.
wellN	The maximum number of wells to include in a overlay. This function will return at maximum wellN wells per strain within an overlay array. These wells are chosen because they have the lowest average object length within the strains included in the overlay array.
proc.img.dir	Supply one of two options. 1: The full path to a directory holding all processed images. For example, "~/proc_images/". 2: A variable name in data that holds the full path to the directory holding processed images matching the data. This option is helpful if the processed images are not kept in the same directory. Users will need to add this variable to the data frame themselves. For example, "image_path". For either option to work the processed images must have the standard _overlay.png suffix and file name output from CellProfiler.
well.label	A variable name in data to display as a well label. For example, "Metadata_Well".
out.dir	The full path to an existing directory that will be used to hold the overlays.

Value

A dataframe nested by the grouping variables with or without flagged objects removed.

AndersenLab/easyXpress documentation built on Dec. 6, 2024, 4:04 a.m.