From 5afe819724d8a08c8ea7f640e41caa5f92020f41 Mon Sep 17 00:00:00 2001
From: Rodrigo Arias Mallo <rodrigo.arias@bsc.es>
Date: Wed, 3 Mar 2021 12:40:15 +0100
Subject: [PATCH] osu: add impi figure

---
 garlic/fig/index.nix  |  1 +
 garlic/fig/osu/impi.R | 67 +++++++++++++++++++++++++++++++++++++++++++
 2 files changed, 68 insertions(+)
 create mode 100644 garlic/fig/osu/impi.R

diff --git a/garlic/fig/index.nix b/garlic/fig/index.nix
index f6709d1..850ba96 100644
--- a/garlic/fig/index.nix
+++ b/garlic/fig/index.nix
@@ -60,6 +60,7 @@ in
 
     bw    = customPlot ./osu/bw.R (ds.osu.bw bw.result);
     bwShm = customPlot ./osu/bw.R (ds.osu.bw bwShm.result);
+    impi  = customPlot ./osu/impi.R (ds.osu.bw impi.result);
   };
 
   # The figures used in the article contained in a directory per figure
diff --git a/garlic/fig/osu/impi.R b/garlic/fig/osu/impi.R
new file mode 100644
index 0000000..e2b7418
--- /dev/null
+++ b/garlic/fig/osu/impi.R
@@ -0,0 +1,67 @@
+library(ggplot2)
+library(dplyr, warn.conflicts = FALSE)
+library(scales)
+library(jsonlite)
+
+args=commandArgs(trailingOnly=TRUE)
+
+# Read the timetable from args[1]
+input_file = "input.json"
+if (length(args)>0) { input_file = args[1] }
+
+# Load the dataset in NDJSON format
+dataset = jsonlite::stream_in(file(input_file), verbose=FALSE) %>%
+	jsonlite::flatten()
+
+# We only need the nblocks and time
+df = select(dataset, config.unitName, config.nodes, config.ntasksPerNode, config.cpusPerTask, config.threshold, size, bw) %>%
+	rename(unitName=config.unitName) %>%
+  rename(threshold=config.threshold)
+
+nodes = unique(df$config.nodes)
+tasksPerNode = unique(df$config.ntasksPerNode)
+cpusPerTask = unique(df$config.cpusPerTask)
+df$unitName = as.factor(df$unitName)
+df$sizeFactor = as.factor(df$size)
+df$threshold = as.factor(df$threshold)
+
+df = group_by(df, unitName, sizeFactor) %>%
+  mutate(medianBw = median(bw)) %>%
+  ungroup()
+
+breaks = 10^(-10:10)
+minor_breaks <- rep(1:9, 21)*(10^rep(-10:10, each=9))
+
+p = ggplot(data=df, aes(x=size, y=bw)) +
+	labs(x="Size (bytes)", y="Bandwidth (MB/s)",
+              title=sprintf("OSU bandwidth benchmark: nodes=%d tasksPerNode=%d cpusPerTask=%d",
+			    nodes, tasksPerNode, cpusPerTask), 
+              subtitle=input_file) +
+	geom_boxplot(aes(color=threshold, group=interaction(threshold, sizeFactor))) +
+	scale_x_continuous(trans=log2_trans()) +
+	#scale_y_log10(breaks = breaks, minor_breaks = minor_breaks) +
+	theme_bw() +
+	theme(legend.position = c(0.8, 0.2))
+
+ppi=300
+h=4
+w=8
+ggsave("boxplot.pdf", plot=p, width=w, height=h, dpi=ppi)
+ggsave("boxplot.png", plot=p, width=w, height=h, dpi=ppi)
+
+p = ggplot(data=df, aes(x=size, y=medianBw)) +
+	labs(x="Size (bytes)", y="Bandwidth (MB/s)",
+              title=sprintf("OSU benchmark: osu_bw",
+			    nodes, tasksPerNode, cpusPerTask), 
+              subtitle=input_file) +
+	geom_line(aes(color=threshold, linetype=threshold)) +
+	geom_point(aes(color=threshold, shape=threshold)) +
+  geom_hline(yintercept = 100e3 / 8, color="red") +
+  annotate("text", x = 8, y = (100e3 / 8) * 0.95, label = "12.5GB/s (100Gb/s)") +
+	scale_x_continuous(trans=log2_trans()) +
+	#scale_y_log10(breaks = breaks, minor_breaks = minor_breaks) +
+	theme_bw() +
+	theme(legend.position = c(0.8, 0.2))
+
+ggsave("median-lines.png", plot=p, width=w, height=h, dpi=ppi)
+ggsave("median-lines.pdf", plot=p, width=w, height=h, dpi=ppi)