NanoStringQC() where the signal to noise ratio input threshold was not being compared against the observed SNR because the parameter name was the same as the column name. Hence the comparison was against itself, and normFlag only checked the percent of genes detected.epiR, forcats, tibble from Imports, moved magrittr to Suggestsnormalize_pools() function to use common pool samples to correct for batch effects
when normalizingnormalize_random() function to use randomly selected samples as the reference for refMethod()fix calculation of genes detected so that the limit of detection is compared in parallel to each sample's counts instead of being recycled, #20
use GitHub Actions for CI, replacing Travis and Appveyor
add gene label "PC_1" for checking smallest positive control
more robust extraction of numeric concentrations from PC gene labels
use RCC file names in parsed data of read_rcc(). Also rename gene name CD3E to CD3e for compatibility purposes
update roxygen
remove Rplots.pdf generated from tests, removed deprecated context()
reduce package dependencies from Imports
remove old packages from Suggests
performance gains in refMethod()
raw parameter in NanoStringQC() now accepts tibbles
check for presence of "Positive" and "Negative" genes in NanoStringQC()
remove most code lints
unit tests for read_rcc() and internal functions check_colnames(), check_genes()
fix bug using NanoStringQC() for single sample data
sort sample names in HLD and OVD cohorts by numeric order, not lexicographic order
tidy up code in HKnorm()
use tidy evaluation in NanostringQC()
use vapply() over sapply() for input checking functions