v0.10.1 metadata and bugfixes

• Package metadata: bioanalyzeR is now covered by the MIT License
• Package metadata: The online vignette and author's email are no longer hosted at stanford.edu
• Package metadata: The png package dependency that was supposed to be removed in v0.10.0 is now actually removed in the package DESCRIPTION
• Bugfix: Problems when reading Bioanalyzer data with missing or conflicting marker peaks
• Bugfix: summarize.custom() skipped rows from bad samples but now outputs NA instead

Sorry it's taken so long to get to some of these, and other Issues. Since the last release I've relocated to another institution and continent, hence the hosting change, so that was keeping me very busy. I think things are back on track now so please do submit new Issues if you have them.

v0.10.0 raw data plots and calibration metadata

• Data structure changed: electrophoresis$mobility.function expanded to electrophoresis$calibration, which includes the same mobility function but also the inverse function and the ladder peak data. Important: data imported with older versions of bioanalyzeR will not be compatible with functions that now require the new $calibration member (currently stdcrv.electrophoresis, qc.electrophoresis, and rawplot.electrophoresis) but may work normally otherwise.
• New feature: rawplot.electrophoresis is a new wrapper for qplot.electrophoresis that generates plots of raw data in the same form as the Agilent software; this is not a good way of viewing data but it is helpful for comparison.
• New feature: read.electrophoresis and similar functions in the family now allow extrapolate = TRUE in case you want to let the mobility model extrapolate outside the range of ladder peaks, but this is likely to be inaccurate since the only good model functions currently use interpolation.
• Old feature removed: TapeStation gel images, previously deprecated, are no longer supported. You must export the raw data in CSV form.
• Old dependency removed: The png package is no longer required now that TapeStation gel images are no longer supported.
• Bugfix: Peak tables exported from ProSize software in the incorrect format are now detected and reported, instead of triggering a vague error message.

v0.9.2 TapeStation bugfix

Bugfix: Unable to read TapeStation data with the "Ladder sizing changed" observation

Documentation: read.electrophoresis help page clarified and reorganized

v0.9.1 ProSize bugfix

Bugfix: In ProSize output, exceptionally bad peaks with NaN percent concentration are now removed like other bad peaks, instead of causing the entire reading operation to fail with the "too many markers" error.

v0.9.0 plot window normalization and filling only marker peaks

• Feature: qplot.electrophoresis now offers the normalize = "window" option, which normalizes with only the data within the x-limits of the plot, as opposed to normalize = "total", which normalizes against the entire sample (between markers). See the updated vignette for example usage.
• Feature: qplot.electrophoresis now offers the show.peaks = "markers" option, which fills in and color-codes the areas of only the marker peaks but not any other peaks. This can be useful if you include the markers in your plot (e.g. raw data electropherogram) and need to explain what they are but aren't otherwise interested in the reported peaks. See the updated vignette for example usage.
• Important syntax change: In qplot.electrophoresis, normalize = TRUE is no longer accepted. That behavior, i.e. normalizing by the entire sample, now requires normalize = "total". Please update old scripts, or consider switching to the new normalize = "window" if that is more useful.
• Important syntax change: In qplot.electrophoresis, show.peaks = FALSE is no longer accepted. That behavior, i.e. not filling in the area under any peaks, now requires show.peaks = "none". Please update old scripts, or consider switching to the new show.peaks = "markers" if that is more useful.
• Syntax improvement: qplot.electrophoresis now uses match.arg for the normalize, geom, and show.peaks arguments so you can simplify them with unambiguous substrings (e.g. normalize = "w" instead of normalize = "window"). It does not use match.arg for the log argument because the options for that are not well suited (copied from ggplot2::qplot).

v0.8.1 minor fixes

• Bugfix: qplot.electrophoresis would override your color or fill argument if facets = NULL
• Bugfix: formatting typo in sample name from demo data used in vignette
• Feature: read.tapestation.csv now gives a warning if it looks like you exported aligned data by mistake (not sure how foolproof this is)
• Feature: added Support section to readme

v0.8.0 violin plots and new TapeStation import

• Important change for all users: The default y-variable in qplot.electrophoresis and default sum-variable in all the integration functions is now "molarity" instead of "concentration", because that seems more intuitive and physically meaningful in most situations. Scripts that previously relied on the default setting will now behave differently.
• Important change for TapeStation users: read.tapestation now accepts an electropherogram CSV file exported from the new version of the TapeStation software instead of the old workaround to extract a gel image screenshot. This is much easier and extracts cleaner data. The old workaround for gel images will be removed in a later release. Many thanks to Agilent for listening to our requests for this feature!
• New feature: Introducing electric violin plots!
• Improvement: The demo data from version 3.2 of the TapeStation software are replaced by the demo data from version 4.1, and instead of gel image PNGs they have CSVs, substantially reducing the size of the demo data.
• Bugfix: calculate.concentration would fail for all samples if any sample was missing its marker peaks.

v0.7.3 bugfix

Bugfix: read.prosize() misidentified the ladder if ProSize-exported data had a well without any peaks

v0.7.2 bugfix

Bug fixed (again): Better and more thorough solution to sporadic problems detecting markers when the true ladder concentrations are known (Bioanalyzer), improving on the patch in v0.7.1.

v0.7.1 bugfix

Bug fixed: Wrong test for detection of ladder peaks when true concentrations are provided (Bioanalyzer), which curiously affected some batches but not all. Also added a stopifnot() to report any further problems at that step.