This directory contains a set of scripts to analyze the simulation results and the radiocarbon dataset. These scripts assume that the simulations have been run according to the script ../neolithic_cpp/simulation_runs.fsh; they can be adapted to other parameter combinations by changing the input files and parameter ranges.
The script c14_data_preprocess_new.r performs a binning on the radiocarbon dataset using the rcarbon package as a preprocessing step.
Inputs:
population/2021-09-08-dataset-nodup-no-sd-filter.csv
Output:
population/c14_data_bins1.csv
Runtime: less than a minute.
The script c14_model_new.py performs the main processing necessary on the radiocarbon dataset. This includes aggregating in space (into overlapping grid cells), calculation of SPDs, fitting of a logistic trend, detrending by generating 1,000 synthetic datasets in each region and identification of ACF minima. The output also includes coefficients of variation and p-values that characterize the deviation from the logistic trend in each region.
This step uses the output of the previous step, along with the grids created during the preprocessing.
Inputs:
population/c14_data_bins1.csv
new_grids_all.csv
Outputs:
population/res_min.csv
population/res_pval.csv
population/res_cv.csv
Runtime: up to 9-10 hours.
The script c14_model_plots1.r reads the results of the previous steps and produces plots of regional time series, ACFs and distribution of ACF minima, CV values and p-values among the regions.
Inputs: output of the previous step
Outputs: panels for Fig. 2 in the main text and SI Figs. S6-S9.
Runtime: a few minutes.
The script [acf_new_aggr_ex.r] creates plots of example time series of simulation results, created with [../neolithic_cpp/simulation_runs.fsh].
Inputs:
simulation/run*spaggr.out.gz
new_grid_ids.csv
new_grid_all.csv
Outputs: top panels for Figs. 4 and 5 in the main text and SI Figs. S10-S13.
Runtime: a few minutes.
The script [acf_new_aggr_plot_rep.r] creates plots of ACF and CV distributions (histograms) from repeated simulation results, in comparison with the radiocarbon data results. This is based on the results created with the script [../neolithic_cpp/simulation_runs_rep.fsh].
Inputs:
results from the previous step
simulation/runs_rep/*RData
population/new_analysis/res_min.csv
population/new_analysis/res_cv.csv
Outputs: bottom panels for Figs. 4 and 5 in the main text and SI Figs. S14-S40.
Runtime: a few minutes.