flEvalF1ScoreFCS: Given a DensityData object with existing thresholds,...
In mlux86/flowLearn: flowLearn

Description Usage Arguments Value Examples

This method gates each cell and compares the result to a true vector of true cell assignments. It returns its performance in the form of various parameters.

1 2	flEvalF1ScoreFCS(densdat, fcs, population, trueAssignments, parentExprs, negate = FALSE)

`densdat`	The full DensityData object with density and densities and set/predicted thresholds.
`fcs`	The name of the FCS file to consider. This is needed to filter the relevant entries in densdat.
`population`	The name of the population to gate. This is needed to filter the relevant entries in densdat.
`trueAssignments`	A vector of length <number of cells>, indicating true cell memberships of the target population.
`parentExprs`	A n*d expression matrix of the parent population where n is the number of cells and d is the number of channels. The channel indices correspond to the channelIdx values in densdat.
`negate`	Whether the gated population should be negated/inverted, i.e. not the cells within the gate but out of it are considered.

A data.frame with precision, recall, f1, trueProportion, predictedProportion.

# In this example, we evaluate true against the true gate assignments, hence it should give perfect performance, i.e. F_1 = 1

f <- unique(flData(flSampleDensdat)$fcs)[[1]]
flEvalF1ScoreFCS(flSampleDensdat, f, 'bcell', flSampleBcellEvaluationData[[f]]$gateAssignments, flSampleBcellEvaluationData[[f]]$parentExprs, FALSE)