Still to do on figures:
Plots with and without skew/kurtosis – DONE, enabled for Figure 1 & 2
Combine plots into full panels – DONE for Figure 1 & Figure 2
separate plotting codes from analysis codes (at least whenever takes any computational effort/time to produce). – DONE
bootstrap_indicators should have option of which indicators are included – DONE
all_indicators should have option of which indicators are included – DONE
make tau_dist_montecarlo generic on type of indicator – DONE
Repeat bootstrap of taus on the CaCO3 and Deut3 with LSN models – RUNNING (figure2.R)
Run montecarlotest of taus on LSN models for CaCO3 and Deut3 – RUNNING (figure3.R)
Guess names for columns in list, if given – DONE
Labels for outer-margins on panel plots – DONE
Adjust bootstrap panel plots to use common y-axis height – DONE
Need simulated data of stable system for Figures 1, 2,3 – DONE
Select empirical-data examples – DONE
Consider empirical-data without any signal?
Considering possible data-sets:
simulated dataset with loss of stability/degenerating environment: LTC example: load(“5550863240.Rdat”), using LSN example now with stable and unstable examples.
simulated dataset without loss of stability, also showing positive signals – DONE
simulated dataset with loss of stability, but showing no signal/negative correlation, not done, not sure if this is necessary to illustrate in Fig 1.
Empirical dataset showing positive correlations which we’ll support in Figure 3, support with considerable uncertainty in Figure 2. Deut3
Empirical dataset showing positive correlations which will appear spurious in Figure 3?
Examples that look at least a little more promising in Figure 2?