Skip to content

v0.9.24 (CRAN)

Compare
Choose a tag to compare
@RLumSK RLumSK released this 07 Jun 14:38
· 787 commits to master since this release
2c5cf8a

Changes in version 0.9.24 (2024-06-07)

This package version requires R >= 4.3

New functions

  • trim_RLum.Data(): This new function enables trimming off the number
    of channels of all supported RLum.Data-class on the time domain. For
    instance, an OSL curve (RLum.Data.Curve-class) has 1000 channels;
    one may want to extract only channels 10 to 100. Moreover, sometimes,
    RLum.Data-class objects have a different number of channels but are
    otherwise measured with the exact time resolution. Until now, it was
    possible to merge those curves, but the subsequent analysis usually
    failed because most of the analysis functions check whether the number
    of channels matches.

  • import_Data(): A convenience wrapper around all functions commencing
    with read_. The functions steps through all functions and the one
    that can import a file is chosen. This function simplifies writing of
    scripts for different data formats.

The Thermochronometry functions (internal)

  • Add internal function to import thermochronometry XLSX sheets into R;
    an R implementation of STAGE1, ExcelToStructure; the function has no
    user visible use for the moment.

Bugfixes and changes

analyse_IRSARRF()

  • The function crashed for RF_nat.lim settings with two parameters;
    fixed (thanks to Mariana Sontag-Gonzalez for reporting)
  • The function gained a new argument option for method: "VSLIDE". If
    set, the vertical sliding with the range set to "auto" is used for
    the estimation of the equivalent dose.

analyse_SAR.CWOSL()

  • The function gained a new argument trim_channels. The default is
    FALSE to do not break existing code. If TRUE OSL and IRSL are
    curves are checked for the lowest number of channels on the curves and
    then all curves are trimmed accordingly using trim_RLum.Data(). This
    should fix an issue where data could not be analysed due to different
    channel numbers erroneously produced by the luminescence readers.
  • The dose points names are returned as factors (which they are)
    instead of characters, with their class in the correct order. If you
    expect character values this may break existing code.
  • The $data element gained two new columns ALQ and POS where the
    functions tries to store data about the running aliquot number and the
    position of the aliquot in the reader. The latter can only be accessed
    if this information was provided in measurement data (e.g., XSYG or
    BIN/BINX)
  • The argument rejection.criteria gained a new option
    recuperation_reference preset to "Natural" to specify the
    regeneration reference dose point taken into account for calculating
    the recuperation rate (suggested by Anna-Maartje Boer and Jakob
    Wallinga)

calc_Huntely2006()

  • If the age is NA due to Ln/Tn values lower than the smallest
    simulated Lx/Tx value, a warning with an explanation is returned.
    Thanks to Christina Neudorf for providing the test dataset.
  • Better catch a few uncontrolled errors and improve the success of the
    fitting (thanks to Salome Oehler for providing scripts and datasets)
  • Regression: The changes in the last version led to unexpected function
    behaviour in cases where users tried the fading correction with only a
    few dose points in the lower part of the dose-response curve because
    the fits were no longer forced through the origin. This issue is now
    corrected, and the user can use the argument
    fit.force_through_origin (which is passed down to
    plot_GrowthCurve()) to force the fit again through the origin
    (thanks to Junjie Zhang for reporting the issue).

convert_Activity2Concentration()

  • The function did not work as expected if input was provided as
    abundance and the output data frame was missing values; fixed.
  • The parameter input_unit now expects either "activity" (the
    default) or "abundance" instead of "Bq/kg" or "ppm/%'. The old
    input parameters are still silently accepted.
  • Potentially breaking change: The naming of the output data frame
    now reports SI unit conform values, i.e. we ppm is no longer
    acceptable because it might have different interpretations. To reflect
    this change, in the function title we replaced “concentration” with
    “abundance”.

read_BIN2R()

  • Argument n.records can now be provided as a vector to specifically
    select records for the import. All other records are skipped and not
    imported. This can be useful in particular for very large
    BIN/BINX-files because allows importing single records very quickly
    without parsing the entire file
  • The function is now less talkative for zero byte records and stopped
    spamming the console.
  • The function learned about RECTYPE 128 records and can now import
    them (before those records were skipped). As it is in the original
    format, these information are only appended to the BIN/BINX file,
    means they do not really represent measurement data. Please note that
    this support is highly experimental!
  • The download of files now uses the new internal function
    .download_file()

read_Daybreak2R()

  • The function gained the ... argument for compatibility reasons

read_RF2R()

  • The function gained the ... argument for compatibility reasons

read_SPE2R()

  • The function gained the ... argument for compatibility reasons
  • The argument verbose was not fully respected; fixed

read_TIFF2R()

  • The function gained the ... argument for compatibility reasons

read_XSYG2R()

  • Tries a little bit harder to determine the correct file path.
  • Returns the file path as @info object in the RLum.Analysis-class
    output.
  • If the file is not readable it only shows the R error and not anymore
    the internal error.

plot_DetPlot()

  • New argument plot: The function gained the new parameter plot to
    provide the possibility to disable the plot output for more complex
    operations
  • Additional input: The function now understands lists of
    RLum.Analysis-class objects and iterates over those objects
    automatically.
  • New argument multicore: As an addition to the list operation, the
    function can let it run in a multi core session
  • The legend displayed the wrong labels for the equivalent dose and the
    Ln/Tn values; fixed.
  • The legend is now disabled by default if no shine-down curve is shown;
    it can be activated again manually using legend.
  • The par settings were not correctly reset; fixed.
  • The function now supports the trim_channel argument from
    analyse_SAR.CWOSL()

plot_GrowthCurve()

  • Edge case: If the Ln/Tn values was way below the fit and
    fit.method = EXP was chosen for the fitting, the function still
    returned a positive equivalent dose for mode = interpolation. Now
    the returned value is NA since no interpolation was possible.
  • Minor graphical polish: the dashed lines are drawn to the end of the
    plot area.

plot_RLum()

  • The function knows how to process results produces by
    analyse_SAR.CWOSL(), analyse_pIRIRSequence() and
    analyse_IRSARRF(); the functionality is very basic and for fast
    reporting situations only.

report_RLum()

  • The file *.Rds file path download did not function anymore; fixed.
  • The function threw an non-conclusive warning; fixed.

Risoe.BINfileData-class()

  • The show method now knows about RECTYPE 128 and will account for
    it.

Risoe.BINFfileData2RLum.Analysis()

  • The function learned about RECTYPE 128 records. Those records
    contain ROI data from camera measurements and are skipped for the
    moment because they do not really belong to measurement data. This may
    change future.

Internals

  • Add internal download helper .download_file() for a more consistent
    approach to detect URL schemes and download files across the package
  • Cover tricky edge case for the internal function
    .expand_parameters(). The function is used to recycle and expand
    function parameters if the input is a list object. It apparently
    showed odd behaviour if a function was used in a loop with objects of
    its own, e.g., f(x = i + 1), then the function would recycle i + 1
    and then evaluate i + 1 in the function environment. Usually, this
    should throw an error, but in particular, i is an object name
    commonly used in loops, so the function went upwards in the
    environments until it found the first i. However, this is not the
    i inserted by the user. The result was that the loop did not seem to
    work for no apparent reason. Now, .expand_parameters() evaluates
    input in the parent environment before passing it down.