Quick summary
Accidentally messed up so the number of replicates for the last two sims are off.
# A tibble: 6 × 7
T_end n keep_rate n_events_mean n_events_sd sim_s_mean fit_s_mean
<dbl> <int> <dbl> <dbl> <dbl> <dbl> <dbl>
1 5 350 0.883 20.4 7.43 0.208 0.636
2 15 350 0.974 69.0 16.0 0.445 1.78
3 25 350 0.98 118. 22.1 0.702 2.96
4 50 350 0.983 244. 30.1 2.20 6.94
5 75 310 0.994 368. 37.7 4.12 11.6
6 100 390 0.995 491. 43.1 6.66 18.5
Cap-Exceeded Replicates
Same as BA, filtering out explosive sims.
df %>%
summarise(
n_total = n(),
n_keep = sum(keep, na.rm = TRUE),
n_dropped_caps = sum(drop_reason == "cap_exceeded", na.rm = TRUE)
)
# A tibble: 1 × 3
n_total n_keep n_dropped_caps
<int> <int> <int>
1 2100 2033 67
df_non_filtered <- df
df <- df %>% filter(keep)
Parameter estimates vs T_end
Dashed line = true value.
plot_estimates_vs_T_ba_bip(df, log_scale = FALSE)
plot_estimates_vs_T_ba_bip(df, log_scale = TRUE)
RMSE decay vs T_end
plot_rmse_vs_T_ba_bip(df)
Keep rate vs T_end
plot_keep_rate_vs_T(df_non_filtered)
