ChemoSpec2Dgain the ability to produce
plotlygraphics, which required changes here in
ChemoSpecUtils. See the news files for
ChemoSpec2Dfor more details.
sumSpectracorrected to show how to pass
ChemoSpeconly at this time.
Col7added to the color options.
rowDist, which supports a number of functions, was overhauled to address confusion in the documentation, and in my head, about distances vs. similarities. Also, different definitions found in the literature were documented more clearly. The Minkowski distance option was removed (ask if you want it back), code was cleaned up, documentation greatly improved, an example was added and unit tests were added. Plot scales were also corrected as necessary. Depending upon which distance option is chosen, this change affects
ChemoSpecas well as
.shrinkLeafnow shrinks the size of leaf labels more aggressively as the number of samples increases.
removeFreq.Spectranow accepts a formula for
sampleDist.Spectramoved here from
ChemoSpecor created anew, with changes to the plot details for consistency between the two classes.
.getLimitsnow knows how to handle negative values in the formulas.
sumSpectra.Spectranow respects/accepts the
...argument, used to pass
tol. This was accidentally disabled.
chkSpectramethods now check for duplicated sample names.
plotScoresnow accepts user supplied values for
ylim; the default calculation is still available.
rowDist(..., method = "cosine")now has a unit test.
dontrunsections in examples, needed for last submission.
getVarExplainedhandles this common task for all data reduction methods.
plotScree.miawas eliminated via the introduction of
.getVarExplainedand small changes to
conColSchemewas moved from
ChemoSpecUtilsso it could be readily available for
Spectra2Dobjects. It was also re-written in the style of
updateGroupsfor consistency and user-friendliness.
colorSymbolwas moved from
plotScoreswas fixed. This was a long standing bug that somehow escaped notice (dating to the first few months of
.getVarExplainedwas returning the eigenvalues, not the percent variance explained, for
class(mia). This affected scree plots for
plotScoreswas no longer including the percent variance explained in score plots, if
ylabwere not provided. Fixed.
chkSpectra.Spectra2Dwould fail if there was only one sample when checking to see if NAs were present in the same positions across all data sets. Fixed.
rowDist. The previous version was giving slightly wrong answers.
updateGroupsnow reports the old groups if
new.grps = NULL, as intended and claimed in the documentation.
plotScreenow knows what to do with class
.chkArgssince errors were introduced via the most recent "fix"! Updated unit tests for the function, but even more work should be done there. Using
HandyStuff::examplePageto look for errors and warnings. None found.
.chkArgs(mode = 13)that was giving an error in
.groupNcolor(the problem seems to have been present from the very first version of
ChemoSpec). Reported by Reinhard Kerschner (many thanks!).
check4Gapsgap checking more robust, with better examples and documentation. If no gap found, a data frame with one row is returned. In turn, this required changes over in
binSpectrawhich are in
ChemoSpec. In addition, while
check4Gapsstill accepts an argument
tolno other functions are using it, and a much more suitable default value is computed in
check4Gapsrather than being computed in another function and passed here.
removeSamplego full S3 dispatch for consistency among the three packages.
ChemoSpecand converted to S3 dispatch.
chkArgsand associated unit tests added, but it is not yet used in any of the other functions.
Seems to work correctly on behalf of
ChemoSpec2D devel versions.
ChemoSpec2Dfor CRAN, temporarily.
.groupNcoloradded with argument mode.
addLegendgets an example.
Any scripts or data that you put into this service are public.
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.