NEWS.md
In wfinsinger/tapas: Trend And PeAkS analysis in paleo-ecological time series

tapas 0.1.5

check_pretreat() function: added some lines to flag for the presence of overlapping sample depths. Modified some other warnings as well.
added the get_overlap_depths() function.

modified the pretreatment() function to calculate sediment-accumulation rates in a different way: now the code uses explicit sample thicknesses and sample-deposition times rather than diff(depth) and diff(age). Many thanks to Pat Bartlein for flagging this issue (26/02/2024).
Added the pretreatment() function from the paleofire v1.2.4 R package (30/10/2023).
Corrected the calculation of peak magnitude (06/06/2023).
Added sample data sets to run examples for arco() function.
Added the arco() function to screen charcoal-area records.
Added two sample data sets (rand_peaks) and (red_noise) for users to play around.
Added the cpts_ar() function that determines zone boundaries for single proxy accumulation-rate records based on a change-point analysis, and checks for the influence of sediment-accumulation rates on the detected change points.

Added the plot_raw() function to produce a polygon plot and optionally overlay that with a bar plot of the raw input data.
Added two functions (tapas2mgcv and mgcv2tapas) that allow determine the ‘background trend’ with a GAM-fitted model. This allows to by-pass the SeriesDetrend() function.

Slightly modified the local_thresh() function. Now it also accommodates discontinuous and/or non-binned data sets. Though, this is currently at a work-in-progress stage, as the SeriesDetrend() function still requires continuous and binned temporal series (i.e. an output from the pretreatment_data() function).
Got rid of trailing horizontal white spaces.
Fixed issue in pretreatment_data() arising when only one variable is included.
Fixed issue with the SNI as it did not call the correct data.
Added a new function export_tapas() that gathers in one data.frame the output of the peak_detection(), local_thresh(), or global_thresh() functions. Many thanks to Kristen Beck for suggesting this.
The selection of values to be evaluated with the local GMMs is now simply based on the “smoothing.yr” parameter. Previously, the selection was based on their index, which was more tedious.
Fixed issue with “Plot.Anomalies()”, as y-axis labels were not printed correctly.