commit 8054411cdd0c1fc01136cb0ecfbd6c7014cdc2bb Author: Karsten Loesing <karsten.loesing@gmx.net> Date: Mon Nov 12 09:58:43 2012 -0500 Handle duplicate values in almost-fast exits graph. These duplicate values came from running the graph data generating script more than once per hour which confused the R script. Fixes #7452. --- rserve/graphs.R | 5 +++-- 1 files changed, 3 insertions(+), 2 deletions(-) diff --git a/rserve/graphs.R b/rserve/graphs.R index 4e61a8d..07de07f 100644 --- a/rserve/graphs.R +++ b/rserve/graphs.R @@ -958,10 +958,11 @@ plot_almost_fast_exits <- function(start, end, path) { t$min_advbw == 2000 & t$same_network == FALSE, ] t <- rbind(data.frame(t1, var = "fast"), data.frame(t2, var = "almost_fast")) - r <- cast(t, valid_after ~ var, value = "relays") + r <- cast(t, valid_after ~ var, value = "relays", fun.aggregate = max) r <- data.frame(valid_after = r$valid_after, fast = r$fast, almost = r$almost_fast - r$fast, var = "relays") - e <- cast(t, valid_after ~ var, value = "exit_prob") + e <- cast(t, valid_after ~ var, value = "exit_prob", + fun.aggregate = max) e <- data.frame(valid_after = e$valid_after, fast = 100 * e$fast, almost = 100 * (e$almost_fast - e$fast), var = "exit_prob") t <- rbind(r, e)