check_gaussians {PrInCE} | R Documentation |
Test whether an input list of Gaussians conforms to the format expected
by PrInCE: that is, a named list with five fields for each entry, i.e., the
number of Gaussians in the mixture model, the r^2 value, the
number of iterations used by nls
, the coefficients of
each model, and the fitted curve.
check_gaussians( gaussians, proteins = NULL, replicate_idx = NULL, n_error = 3, pct_warning = 0.1 )
gaussians |
the list of Gaussians |
proteins |
the complete set of input proteins |
replicate_idx |
the replicate being analyzed, if input proteins are provided; used to throw more informative error messages |
n_error |
minimum number of proteins that can have fitted Gaussians without throwing an error |
pct_warning |
minimum fraction of proteins that can have fitted Gaussians without giving a warning |
Optionally, some extra checks will be
done on the fraction of proteins in the complete dataset for which a
Gaussian mixture model could be fit, if provided. In particular, the function
will throw an error if fewer than n_error
proteins have a fitted
Gaussian, and emit a warning if fewer than pct_warning
do.
TRUE
if all conditions are met, but throws an error if any
is not
data(scott_gaussians) check_gaussians(scott_gaussians)