NEWS.md

Version 1.5.1 (in development)

• Bug Fix: Removed non-relevant warning messages from calibrate() within modelTest().
• NEW: qCal() function computes quantiles of calibrated dates.
• UPDATE: multiplot() now enables relative positioning of labels based on quantiles of calibrated dates and/or a user-defined offset.

Version 1.5.0 (21 November 2022)

• UPDATE: Functions related to spatial permutation test (spweights(),sptest(), and plot.spatialTest()) now depend on sf instead of sp package.
• UPDATE: modelTest() now issues a warning message when model is set to "exponential" or "uniform" (i.e. when the SPD is tested against exponential and uniform curves fitted on observed SPD values.)
• UPDATE: Updated vignette with minor changes reflecting the workflow for sptest() and the warning in modelTest().
• Bug Fix: Fixed a problem with plot.stackCalSPD() function when attempting to display proportion SPD plots including intervals with no dates.
• Updated spatstat dependencies.

Version 1.4.4 (21 July 2022)

• Bug Fix: Fixed a bug in spd() reducing the timeRange when datenormalised was set to TRUE.
• UPDATE: calibrate() now issues a warning in case dates have calibrated probabilities outside the user defined timeRange value.
• UPDATE: stackspd() now requires the group argument to be factors. Factor levels will be condition the plot order in plot.stackCalSPD().
• UPDATE: when setting the rescale=TRUE and type='multipanel', plot.stackCalSPD() will now display the true y-axis range of individual SPDs (instead of a rescaled 0-1 range).

Version 1.4.3 (17 February 2022)

• UPDATE: Reversed legend item order in plot.stackCalSPD() to match display
• UPDATE: hpdi() now has the option to return a character vector with HPDI ranges in either BP or BCAD.

Version 1.4.2 (15 March 2021)

• Bug fix: Fixed a number of minor bugs and errors caused when the summed probability vector contains zero. Functions affected by this bugs were:
  • proportion plots in plot.stackCalSPD()
  • summary() and plot() functions for SpdModelTest class objects
• Bug Fix: Fixed issues with summary() function on CalDates class object with multiple dates (thanks to nferebau for the bug report and fix) plot.stackCalSPD() proportion plot to handle instances with 0 summed probability in the time Range of analysis.
• UPDATE: binMed() function can now handle larger caldates objects.
• NEW: poolDates() function allows combination of 14C ages associated with the same event using Ward and Wilson 1978 method.
• Minor bug fixes.
• Minor typos and errors in the help documentation.
• Updated dependencies for spatstat(>= 2.0-0).

Version 1.4.1 (6 October 2020)

• UPDATE: Improved performance of the calibrate() function (c.a 400% faster)
• UPDATE: multiplot allows for rescaled calibrated probability distribution (new argument rescale) for improved readability.
• UPDATE: The package vignette now includes examples from new functions such as multiplot() and stackspd().
• UPDATE: hdmi() now computes the probability mass of each age bracket (thanks to hanecakr for the suggestion and sample code)
• NEW: transformSPD() allows for taphonomic (and other user-defined) corrections to SPDs.
• Bug fix: Fixed color matching and improved label placing in the multiplot function.
• Bug fix: calCurve='normal' no longer causing error in calibrate()

Version 1.4.0 (15 August 2020)

• IntCal20, ShCal20, and Marine20 curves added, with IntCal20 as default calibration curve for calibrate().

Version 1.3.3 (4 August 2020)

• Minor changes on NAMESPACE and test environment to ensure workable CRAN checks for all operating systems.

Verson 1.3.2 (25 July 2020)

• Bug fix : F14C=TRUE was producting an error message in calibrate() after refactoring in version 1.3.1.
• Bug fix : fixed error in the handling of resErrors (Delta R error) for marine reservoir effect.
• NEW: stackspd() creates a set of multiple SPDs as an object of class stackCalSPD with a dedicated plot function.
• NEW: multiplot() function for displaying multiple calibrated dates.
• NEW: hpdi() function for computing highest probability density intervals of calibrated dates.
• NEW: combine() function for concatenating multiple calDates class objects.
• UPDATE: calibrate(), modelTest(), and sptest() are now parallelised using the doSNOW package, enabling progress bar when running over multiple cores.

Version 1.3.1 (18 March 2020)

• binPrep() now accepts alternative clustering algorithms
• Refactoring of calibration related functions to increase readability
• Minor bug fixes

Version 1.3.0 (12 December 2019)

• UPDATEs in modelTest():
  • Added an error message and a note in the help documentation warning users to not use calibration curves not supported by uncalibrate().
  • NEW argument gridclip to add range constrain in the simulation envelope. When set to TRUE the sampling is executed within the observed range of dates.
  • UPDATE timeRange can now be left undefined, in which case the range of median calibrated observed dates is used.
  • NEW argument normalised handles whether simulated dates should be normalised or not.
  • NEW arguments backsight and changexpr enabling the comparison of expected and observed rates of changes.
  • UPDATE: improved performance of calsample and uncalsample methods (update in uncalibrate()) `
  • Bug fix: users no longer need to specify nsim when fitonly is set to TRUE.
• UPDATE in permTest()
  • NEW arguments backsight and changexpr enabling the comparison of expected and observed rates of changes.
• UPDATE in spd(): aggregation matrix now spans beyond timeRange to avoid edge effects.
• UPDATE in calibrate(): the new argument F14C enables calibration in F14C space.
• UPDATE in SPpermTest():
  • function is deprecated and renamed sptest()
  • function now allows user defined expression for computing the rate of change
• UPDATE in spd2gg():
  • function is deprecated and renamed spd2rc()
  • function now allows user defined expression for computing the rate of change
  • rates of changes can now be based on abutting time-blocks (e.g. 5600-5501 to 5500-5401) or a fixed interval across all years.
• NEW: thinDates() function to randomly select a maximum number of dates per site, bin or phase.
• NEW: as.CalDates() function converts dates in BchronCalibratedDates class (Bchron) and oxcAARCalibratedDatesListclass (oxAAR) into CalDates` class (rcarbon).
• NEW: sampleDates() function for sampling random dates from calibrated dates or bins.
• NEW: ckde() function for generating composite kernel density estimates.
• NEW: Suite of functions for mapping the spatial intensity of a set of radiocarbon dates via Kernel Density Estimates.
  • stkde() Map the spatio-temporal intensity of a set of radiocarbon dates across multiple years
  • spkde() Map the spatial intensity of a set of radiocarbon dates for a given focal year.
  • Associated plot() function.
• NEW: subset() and which.CalDates() for subsetting and extracting indices of calibrated dates based on temporal intervals described by logical conditions (e.g. between 5500 and 4500 cal BP) and user defined probability mass.
• Bug fix: fixed display error in some plot() functions when calendar is set to `BC/AD'
• Bug fix: fixed error in the computation of the combined uncertainties in mixCurves().
• Further minor bug fixes for R 4.0.0

Version 1.2.0 (1 October 2018)

• Bug fix in SPpermTest(): Functions was not working when not running when raw=FALSE and ncores=1.
• Bug fix in summary() for the output of modelTest(): The number of bins reported was incorrect.
• UPDATE: binsense() requires a smaller number of non-optional arguments and allows for binning based on median calibrated dates.
• UPDATE: plot() function for geometric growth rates now allows for BC/AD calendar display.
• UPDATE: binPrep() can now group dates based on median calibrated dates.
• UPDATE: modelTest()can now test SPD generated from multiple calibration curves.
• UPDATE: modelTest() now uses randomised back-calibration for the 'calsample' method.
• UPDATE: uncalibrate()'s random back-calibration now utilises both the user supplied lab error and the error of the calibration curve.
• NEW: mixCurves() function generates mixed terrestrial/marine calibration curves.
• Further minor updates in plot labels, help documentation, and vignette.

Version 1.1.3 (26 June 2018)

• Bug fix in permTest(): p-values calculation algorithm was yielding p-values of 0.
• NEW: summary() functions for the outputs of permTest() and modelTest().

Version 1.1.2 (22 May 2018)

• Bug Fix in calibrate(): calibration of multiple dates with different calibration curves was based on the calibration curve of the first sample.
• Bug Fix in modelTest(): back-calibration routine was ignoring user-supplied calibration curve and was solely using "intcal13".
• UPDATE: modelTest() now allows for two distinct procedures for generating random dates from fitted models.
• Minor corrections (e.g. typos) in the help documentation.

Version 1.1.1 (28 April 2018)

• Added a Vignette.
• Added normally-distributed (non-14C) age in calibrate().
• Fixed a bug in SPpermTest() generating opposite results (positive deviations were recorded as negative deviations) when ncores was larger than 1.
• Fixed a minor bug in the implementation of the [North el al 2002] formula
• Several changes in theplot.SpatialTest().
• NEW function: spd2gg() to convert SPD curves into geometric growth rates for given temporal blocks and an associated plot function.
• Minor updates in the documentation.

Version 1.1.0 (12 March 2018)

• Improved performance of modelTest() function when running with multiple cores.
• NEW function: p2pTest() for comparing point to point differences in SPD.
• p-values of Monte-Carlo simulations are now all calculated using a different formula.
• minor bug fixes

Version 1.0.0 (9 October 2017)

First official CRAN release.