- Added an example for reservoir correction in the vignette
- Fixed an example in Bchronology that was adding titles/labels incorrectly
- Updated vignette to show how to customise plots
- Fixed a bug where an error for identical positions (with no thickness errors) was not being shown
- Fixed an issue with positionNormalise
- Fixed an issue with jitterPositions by replacing it with artificialThickness that enables customised amounts of jittering when there are identical positions and no thicknesses
- Forced ids argument in BchronCalibrate and Bchronology to be unique otherwise causes problems with plots
- Added in a BchronCheck function to allow for checking of all argument types using checkmate
- Added in positionEps argument to Bchronology to avoid positions being simulated too close to each other and leading to numerical underflow errors
- Changed default colours of dates in plot.CalibratedDates and plot.BchronologyRun
- Added an extra check that all ageSds are bigger than zero
- Allowed the option for calibrating dates out of calibration curve range
- Fixed bug that occurred with large 14C values close to upper range of calibration curve
- Changed the means by which starting values are created. This was to fix a bug (issue #17 on GitHub) where a middle extreme date (e.g. non-14C) could cause the starting values of the stratigraphically ordered dates to fall beyond the range of the calibration curve.
- Included option to allow bespoke starting values for complex cores
- Changed some of the algorithm details to avoid certain bugs with numerical underflow problems (tested in test_all_chrons)
- Updated the BchronDensity and BchronDensityFast functions to show lines on top of dates
- Included ability in above functions to change amount of date transparency
- Used styler to make the syntax follow tidyverse style guidelines
- Fixed bug in starting values which occasionally produced values before the extractDate
- Fixed bug in choosePositions where position chosen was at the edge of the domain
- Fixed a weird bug where some hdrs were coming out the wrong way in plot.Bchronology
- Add the ability to create OxCal style plots with the calibration curve
- Added a feature to plot individual dates in plot.BchronCalibratedDates
- Included intcal20, shcal20, and marine20 new calibration curves
- Fixed a bug in Bchronology that caused incorrect predictions
- Fixed a(nother) small bug in predict.Bchronologies
- Included proper testthat, travis, and code coverage
- Fixed bug in predict.Bchronologies
Fixed two important bugs:
- Fixed a mistake in BchronCalibrate that will cause calibrated to have incorrect error ranges compared to OxCal
- Fixed a problem that occurred in very straight chronologies where a certain parameter would go to zero and cause numerical underflow
- Fixed some minor issues with ggplots
- Added in code for scaling the x-axis
- Changed plots to use ggplot2 via ggridges
Added some new functions:
- dateInfluence, which works out how influential a single date is in a core
- coreInfluence, which tries to estimate the total uncertainty reduction/increase between two Bchronology runs
- choosePositions, which suggest new positions/depths to date to maximally reduce uncertainty