Chart consistency + visual cleanup

This commit is contained in:
2026-07-26 23:04:36 +02:00
parent df83623101
commit aad3b85d8f
10 changed files with 326 additions and 293 deletions
@@ -1,8 +1,11 @@
library(ggplot2)
library(dplyr)
library(readr)
library(stringr)
# Usage: Rscript combined_comparison.r exp_abspath1 exp_abspath2 ... [resultsdata_file]
# One coordinate system per base experiment (facet); c/aot/interp variants
# share each facet, coloured by variant.
args <- commandArgs(trailingOnly = TRUE)
if (length(args) < 1) {
@@ -20,17 +23,36 @@ exp_args <- if (grepl("\\.csv$", args[length(args)])) {
args
}
extract_info <- function(path) {
dir_name <- basename(path)
match <- str_match(
dir_name,
"^\\d{2}-\\d{2}_\\d{2}-\\d{2}-\\d{2}_(.+?)_(c|aot|interp)_"
)
if (is.na(match[1, 1])) {
warning(paste("Could not parse:", dir_name))
return(NULL)
}
list(base_name = match[1, 2], variant = match[1, 3], path = path)
}
all_data <- data.frame()
for (arg in exp_args) {
csv_file <- file.path(arg, csv_suffix)
info <- extract_info(arg)
if (is.null(info)) {
next
}
csv_file <- file.path(info$path, csv_suffix)
if (!file.exists(csv_file)) {
warning(paste("Missing:", csv_file))
next
}
df <- read_csv(csv_file, col_types = cols())
df$experiment <- basename(arg)
df$base_name <- info$base_name
df$variant <- info$variant
all_data <- bind_rows(all_data, df)
}
@@ -39,7 +61,7 @@ if (nrow(all_data) == 0) {
}
totals <- all_data |>
group_by(experiment, resulttype) |>
group_by(base_name, variant, resulttype) |>
summarise(faults = sum(faults, na.rm = TRUE), .groups = "drop") |>
ungroup()
@@ -57,21 +79,25 @@ totals$resulttype <- factor(totals$resulttype, levels = marker_order)
plot <- ggplot(
totals,
aes(x = resulttype, y = faults, colour = experiment, group = experiment)
aes(x = resulttype, y = faults, colour = variant, group = variant)
) +
geom_point(size = 2) +
geom_line() +
facet_wrap(~base_name) +
scale_y_log10() +
labs(
x = "Marker",
y = "Faults",
title = "Combined Comparison",
color = "Experiment"
x = "Fault Type",
y = "Fault Count",
title = "Fault Count Comparison",
color = "Variant"
) +
theme_minimal() +
theme(axis.text.x = element_text(angle = 45, hjust = 1))
theme(
axis.text.x = element_text(angle = 90, hjust = 1),
plot.title = element_text(size = 14, face = "bold")
)
suffix <- gsub("^resultsdata|\\.csv$", "", csv_suffix)
outfile <- paste0("injections/combined_comparison", suffix, ".svg")
outfile <- paste0("injections/fault_count_comparison", suffix, ".svg")
ggsave(outfile, plot = plot, width = 12, height = 6)
print(paste("Saved", outfile))
@@ -4,8 +4,8 @@ library(readr)
library(stringr)
library(tidyr)
# Usage: Rscript ratio_correlation.r exp_abspath1 exp_abspath2 ... [resultsdata_file]
# Plots correlation between aot/c and interp/c ratios
# Usage: Rscript combined_fault_correlation.r exp_abspath1 exp_abspath2 ... [resultsdata_file]
# Plots correlation between raw aot and interp fault counts (no C baseline).
args <- commandArgs(trailingOnly = TRUE)
if (length(args) < 2) {
@@ -69,48 +69,52 @@ all_data <- all_data |>
group_by(base_name, variant, benchmark, resulttype) |>
summarise(faults = sum(faults), .groups = "drop")
baseline <- all_data |> filter(variant == "c")
comparisons <- all_data |> filter(variant != "c")
# Only aot/interp matter; C is not used as a baseline here.
counts <- all_data |> filter(variant %in% c("aot", "interp"))
ratios <- comparisons |>
left_join(
baseline |> select(base_name, benchmark, resulttype, faults),
by = c("base_name", "benchmark", "resulttype"),
suffix = c("", "_baseline")
) |>
filter(!is.na(faults_baseline), faults_baseline > 0) |>
mutate(ratio = faults / faults_baseline)
if (nrow(ratios) == 0) {
stop("No ratios computed (missing baseline or zero values)")
}
# Pivot to get aot and interp ratios side by side
ratio_wide <- ratios |>
select(base_name, benchmark, resulttype, variant, ratio) |>
pivot_wider(names_from = variant, values_from = ratio) |>
# Pivot to get aot and interp fault counts side by side
counts_wide <- counts |>
select(base_name, benchmark, resulttype, variant, faults) |>
pivot_wider(names_from = variant, values_from = faults) |>
filter(!is.na(aot), !is.na(interp))
if (nrow(ratio_wide) == 0) {
stop("No paired aot/interp ratios found")
if (nrow(counts_wide) == 0) {
stop("No paired aot/interp fault counts found")
}
# Compute correlation
cor_result <- cor(ratio_wide$aot, ratio_wide$interp, method = "pearson")
cat(sprintf("Pearson correlation: %.4f\n", cor_result))
cor_raw <- cor(counts_wide$aot, counts_wide$interp, method = "pearson")
cor_log <- cor(
log10(counts_wide$aot),
log10(counts_wide$interp),
method = "pearson"
)
cat(sprintf("Pearson correlation (raw): %.4f\n", cor_raw))
cat(sprintf("Pearson correlation (log10): %.4f\n", cor_log))
# Create plot
plot <- ggplot(
ratio_wide,
counts_wide,
aes(x = aot, y = interp, color = base_name, shape = resulttype)
) +
# geom_abline(
# slope = 1,
# intercept = 0,
# colour = "grey70",
# linetype = "dotted"
# ) +
geom_point(size = 3, alpha = 0.7) +
scale_x_log10(name = "AOT / C Ratio") +
scale_y_log10(name = "Interpreter / C Ratio") +
scale_x_log10(name = "AOT Fault Count") +
scale_y_log10(name = "Interpreter Fault Count") +
labs(
title = sprintf("Ratio Correlation (r = %.4f)", cor_result),
# title = sprintf(
# "Fault Count Correlation (r_raw = %.4f, r_log = %.4f)",
# cor_raw,
# cor_log
# ),
title = "Fault Count Correlation",
color = "Experiment",
shape = "Marker"
shape = "Fault Type"
) +
theme_minimal() +
theme(
@@ -118,5 +122,7 @@ plot <- ggplot(
plot.title = element_text(size = 14, face = "bold")
)
ggsave("injections/ratio_correlation.svg", plot = plot, width = 10, height = 8)
print("Saved ratio_correlation.svg")
suffix <- gsub("^resultsdata|\\.csv$", "", csv_suffix)
outfile <- paste0("injections/fault_count_correlation", suffix, ".svg")
ggsave(outfile, plot = plot, width = 10, height = 8)
print(paste("Saved", outfile))
@@ -0,0 +1,92 @@
library(ggplot2)
library(dplyr)
library(readr)
# Usage: Rscript combined_fault_rates.r exp_abspath1 exp_abspath2 ... [faults_file]
args <- commandArgs(trailingOnly = TRUE)
if (length(args) < 1) {
stop("Need at least 1 experiment")
}
csv_suffix <- if (grepl("\\.csv$", args[length(args)])) {
args[length(args)]
} else {
"faults.csv"
}
exp_args <- if (grepl("\\.csv$", args[length(args)])) {
args[-length(args)]
} else {
args
}
# Faults / instruction count, per experiment
rates <- data.frame()
for (arg in exp_args) {
faults_file <- file.path(arg, csv_suffix)
mnem_file <- file.path(arg, "mnemonics.csv")
if (!file.exists(faults_file)) {
warning(paste("Missing:", faults_file))
next
}
if (!file.exists(mnem_file)) {
warning(paste("Missing:", mnem_file))
next
}
df <- read_csv(faults_file, col_types = cols())
mdf <- read_csv(mnem_file, col_types = cols())
total_faults <- df |>
filter(resulttype != "OK_MARKER") |>
summarise(faults = sum(faults, na.rm = TRUE)) |>
pull(faults)
total_instrs <- sum(mdf$count, na.rm = TRUE)
if (is.na(total_instrs) || total_instrs == 0) {
warning(paste("Zero instruction count for", arg))
next
}
rates <- bind_rows(
rates,
data.frame(
experiment = basename(arg),
fault_rate = total_faults / total_instrs
)
)
}
if (nrow(rates) == 0) {
stop("No data loaded")
}
# Order by fault rate
rates <- rates |>
arrange(desc(fault_rate)) |>
mutate(experiment = factor(experiment, levels = experiment))
plot <- ggplot(
rates,
aes(x = experiment, y = fault_rate, fill = experiment)
) +
geom_col() +
labs(
title = "Fault Rate per Instruction",
x = "Experiment",
y = "Faults / Instruction Count"
) +
theme_minimal() +
theme(
axis.text.x = element_text(angle = 90, hjust = 1),
legend.position = "none",
plot.title = element_text(size = 14, face = "bold")
)
suffix <- gsub("^faults|\\.csv$", "", csv_suffix)
outfile <- paste0("injections/fault_rates_per_instruction", suffix, ".svg")
ggsave(outfile, plot = plot, width = 12, height = 6)
print(paste("Saved", outfile))
@@ -110,14 +110,15 @@ plot <- ggplot(
vjust = -0.8,
check_overlap = TRUE
) +
scale_x_log10(name = "Instruction Execution Frequency") +
scale_y_log10(name = "Failure-Marker Occurrences") +
scale_x_log10(name = "Instruction Executions") +
scale_y_log10(name = "Fault Count") +
labs(
title = sprintf(
"Instruction / Fault Correlation (r_raw = %.4f, r_log = %.4f)",
cor_raw,
cor_log
),
# title = sprintf(
# "Instruction / Fault Correlation (r_raw = %.4f, r_log = %.4f)",
# cor_raw,
# cor_log
# ),
title = "Instruction / Fault Correlation",
colour = "Mnemonic"
) +
theme_minimal() +
@@ -0,0 +1,120 @@
library(ggplot2)
library(dplyr)
library(readr)
library(viridisLite)
# Usage: Rscript combined_instr_fault_correlation_heatmap.r exp_abspath1 ... [faults_file]
args <- commandArgs(trailingOnly = TRUE)
if (length(args) < 1) {
stop("Need at least 1 experiment")
}
csv_suffix <- if (grepl("\\.csv$", args[length(args)])) {
args[length(args)]
} else {
"faults.csv"
}
exp_args <- if (grepl("\\.csv$", args[length(args)])) {
args[-length(args)]
} else {
args
}
all_data <- data.frame()
all_mnem <- data.frame()
for (arg in exp_args) {
csv_file <- file.path(arg, csv_suffix)
if (!file.exists(csv_file)) {
warning(paste("Missing:", csv_file))
next
}
df <- read_csv(csv_file, col_types = cols())
df$experiment <- basename(arg)
all_data <- bind_rows(all_data, df)
# TODO: This is ignoring any filters currently
mnem_file <- file.path(arg, "mnemonics.csv")
if (!file.exists(mnem_file)) {
warning(paste("Missing:", mnem_file))
next
}
mdf <- read_csv(mnem_file, col_types = cols())
all_mnem <- bind_rows(all_mnem, mdf)
}
if (nrow(all_data) == 0) {
stop("No faults.csv data loaded")
}
if (nrow(all_mnem) == 0) {
stop("No mnemonics.csv data loaded")
}
# no OK_MARKER, sum GROUP1 + TRAP.
all_data <- all_data |>
filter(!is.na(mnemonic), mnemonic != "NULL") |>
filter(resulttype != "OK_MARKER") |>
mutate(
resulttype = ifelse(resulttype == "GROUP1_MARKER", "TRAP", resulttype)
)
if (nrow(all_data) == 0) {
stop("No failure-marker data to plot")
}
# Sum faults per (mnemonic, marker) pair for all experiments
heat <- all_data |>
group_by(mnemonic, resulttype) |>
summarise(faults = sum(faults, na.rm = TRUE), .groups = "drop")
# Sum mnemonic counts for all experiments
mnem_counts <- all_mnem |>
filter(!is.na(mnemonic), mnemonic != "NULL") |>
group_by(mnemonic) |>
summarise(count = sum(count, na.rm = TRUE), .groups = "drop")
# Normalize by mnemonic count
heat <- heat |>
left_join(mnem_counts, by = "mnemonic") |>
filter(!is.na(count), count > 0) |>
mutate(fault_rate = faults / count)
if (nrow(heat) == 0) {
stop("Heat join failed")
}
# Order by fault rate
mnem_order <- heat |>
group_by(mnemonic) |>
summarise(total = sum(fault_rate), .groups = "drop") |>
arrange(desc(total)) |>
pull(mnemonic)
heat <- heat |>
mutate(mnemonic = factor(mnemonic, levels = mnem_order))
plot <- ggplot(
heat,
aes(x = mnemonic, y = resulttype, fill = fault_rate)
) +
geom_tile(colour = "white") +
scale_fill_viridis_c(name = "Fault rate", trans = "log10") +
labs(
title = "Instruction / Fault Rate Heatmap (Normalized)",
x = "Instruction",
y = "Fault Type"
) +
theme_minimal() +
theme(
axis.text.x = element_text(angle = 90, hjust = 1),
panel.grid = element_blank(),
plot.title = element_text(size = 14, face = "bold")
)
suffix <- gsub("^faults|\\.csv$", "", csv_suffix)
outfile <- paste0("injections/instr_fault_rate_heatmap", suffix, ".svg")
ggsave(outfile, plot = plot, width = 12, height = 6)
print(paste("Saved", outfile))
+10 -4
View File
@@ -83,12 +83,18 @@ plot <- ggplot(
) +
geom_point(size = 2) +
geom_line() +
facet_wrap(~base_name, scales = "free_x") +
facet_wrap(~base_name) +
scale_y_log10(name = "Ratio (to C)") +
scale_x_discrete(name = "Marker") +
labs(color = "Variant") +
scale_x_discrete(name = "Fault Type") +
labs(
title = "Fault Count Ratios",
color = "Variant"
) +
theme_minimal() +
theme(axis.text.x = element_text(angle = 45, hjust = 1))
theme(
axis.text.x = element_text(angle = 90, hjust = 1),
plot.title = element_text(size = 14, face = "bold")
)
ggsave("injections/ratio_comparison.svg", plot = plot, width = 12, height = 8)
print("Saved ratio_comparison.svg")
@@ -3,8 +3,8 @@ library(dplyr)
library(readr)
library(stringr)
# Usage: Rscript ratio_comparison_merged.r exp_abspath1 exp_abspath2 ... [resultsdata_file]
# Sums all benchmarks
# Usage: Rscript ratio_comparison_merged_trap.r exp_abspath1 exp_abspath2 ... [resultsdata_file]
# Sums all benchmarks, merges GROUP1_MARKER into TRAP
args <- commandArgs(trailingOnly = TRUE)
if (length(args) < 2) {
@@ -59,7 +59,9 @@ if (nrow(all_data) == 0) {
stop("No data loaded")
}
# Add all benchs together (per marker type)
all_data <- all_data |>
mutate(resulttype = ifelse(resulttype == "GROUP1_MARKER", "TRAP", resulttype))
merged_data <- all_data |>
group_by(base_name, variant, resulttype) |>
summarise(faults = sum(faults), .groups = "drop")
@@ -88,15 +90,21 @@ plot <- ggplot(
geom_line() +
facet_wrap(~base_name) +
scale_y_log10(name = "Ratio (to C)") +
scale_x_discrete(name = "Marker") +
labs(color = "Variant") +
scale_x_discrete(name = "Fault Type") +
labs(
color = "Variant",
title = "Fault Count Ratios"
) +
theme_minimal() +
theme(axis.text.x = element_text(angle = 45, hjust = 1))
theme(
axis.text.x = element_text(angle = 90, hjust = 1),
plot.title = element_text(size = 14, face = "bold")
)
ggsave(
"injections/ratio_comparison_merged.svg",
"injections/ratio_comparison_merged_trap.svg",
plot = plot,
width = 12,
height = 8
)
print("Saved ratio_comparison_merged.svg")
print("Saved ratio_comparison_merged_trap.svg")
@@ -1,100 +0,0 @@
library(ggplot2)
library(dplyr)
library(readr)
library(stringr)
# Usage: Rscript ratio_comparison_merged_trap.r exp_abspath1 exp_abspath2 ... [resultsdata_file]
# Sums all benchmarks, merges GROUP1_MARKER into TRAP
args <- commandArgs(trailingOnly = TRUE)
if (length(args) < 2) {
stop("Need at least 2 experiments")
}
csv_suffix <- if (grepl("\\.csv$", args[length(args)])) {
args[length(args)]
} else {
"resultsdata.csv"
}
exp_args <- if (grepl("\\.csv$", args[length(args)])) args[-length(args)] else args
extract_info <- function(path) {
dir_name <- basename(path)
match <- str_match(
dir_name,
"^\\d{2}-\\d{2}_\\d{2}-\\d{2}-\\d{2}_(.+?)_(c|aot|interp)_"
)
if (is.na(match[1, 1])) {
warning(paste("Could not parse:", dir_name))
return(NULL)
}
list(base_name = match[1, 2], variant = match[1, 3], path = path)
}
all_data <- data.frame()
for (arg in exp_args) {
info <- extract_info(arg)
if (is.null(info)) {
next
}
csv_file <- file.path(info$path, csv_suffix)
if (!file.exists(csv_file)) {
warning(paste("Missing:", csv_file))
next
}
df <- read_csv(csv_file, col_types = cols())
df$base_name <- info$base_name
df$variant <- info$variant
all_data <- bind_rows(all_data, df)
}
if (nrow(all_data) == 0) {
stop("No data loaded")
}
all_data <- all_data |>
mutate(resulttype = ifelse(resulttype == "GROUP1_MARKER", "TRAP", resulttype))
merged_data <- all_data |>
group_by(base_name, variant, resulttype) |>
summarise(faults = sum(faults), .groups = "drop")
baseline <- merged_data |> filter(variant == "c")
comparisons <- merged_data |> filter(variant != "c")
ratios <- comparisons |>
left_join(
baseline |> select(base_name, resulttype, faults),
by = c("base_name", "resulttype"),
suffix = c("", "_baseline")
) |>
filter(!is.na(faults_baseline), faults_baseline > 0) |>
mutate(ratio = faults / faults_baseline)
if (nrow(ratios) == 0) {
stop("No ratios computed (missing baseline or zero values)")
}
plot <- ggplot(
ratios,
aes(x = resulttype, y = ratio, color = variant, group = variant)
) +
geom_point(size = 2) +
geom_line() +
facet_wrap(~base_name) +
scale_y_log10(name = "Ratio (to C)") +
scale_x_discrete(name = "Marker") +
labs(color = "Variant") +
theme_minimal() +
theme(axis.text.x = element_text(angle = 45, hjust = 1))
ggsave(
"injections/ratio_comparison_merged_trap.svg",
plot = plot,
width = 12,
height = 8
)
print("Saved ratio_comparison_merged_trap.svg")
@@ -1,129 +0,0 @@
library(ggplot2)
library(dplyr)
library(readr)
library(stringr)
library(tidyr)
# Usage: Rscript ratio_correlation_no_mem.r exp_abspath1 exp_abspath2 ... [resultsdata_file]
# Plots correlation between aot/c and interp/c ratios, ignoring mem benchmark
# NOTE: Just a copy of the ratio_correlation.r script where I've changed the filter in line 67
args <- commandArgs(trailingOnly = TRUE)
if (length(args) < 2) {
stop("Need at least 2 experiments")
}
csv_suffix <- if (grepl("\\.csv$", args[length(args)])) {
args[length(args)]
} else {
"resultsdata.csv"
}
exp_args <- if (grepl("\\.csv$", args[length(args)])) {
args[-length(args)]
} else {
args
}
extract_info <- function(path) {
dir_name <- basename(path)
match <- str_match(
dir_name,
"^\\d{2}-\\d{2}_\\d{2}-\\d{2}-\\d{2}_(.+?)_(c|aot|interp)_"
)
if (is.na(match[1, 1])) {
warning(paste("Could not parse:", dir_name))
return(NULL)
}
list(base_name = match[1, 2], variant = match[1, 3], path = path)
}
all_data <- data.frame()
for (arg in exp_args) {
info <- extract_info(arg)
if (is.null(info)) {
next
}
csv_file <- file.path(info$path, csv_suffix)
if (!file.exists(csv_file)) {
warning(paste("Missing:", csv_file))
next
}
df <- read_csv(csv_file, col_types = cols())
df$base_name <- info$base_name
df$variant <- info$variant
all_data <- bind_rows(all_data, df)
}
if (nrow(all_data) == 0) {
stop("No data loaded")
}
# Ignore OK_MARKER (only plot failures) and sum GROUP1_MARKER with TRAP
# Also ignore mem benchmark
all_data <- all_data |>
filter(resulttype %in% c("TRAP", "GROUP1_MARKER"), benchmark == "ip") |>
mutate(resulttype = ifelse(resulttype == "GROUP1_MARKER", "TRAP", resulttype))
all_data <- all_data |>
group_by(base_name, variant, benchmark, resulttype) |>
summarise(faults = sum(faults), .groups = "drop")
baseline <- all_data |> filter(variant == "c")
comparisons <- all_data |> filter(variant != "c")
ratios <- comparisons |>
left_join(
baseline |> select(base_name, benchmark, resulttype, faults),
by = c("base_name", "benchmark", "resulttype"),
suffix = c("", "_baseline")
) |>
filter(!is.na(faults_baseline), faults_baseline > 0) |>
mutate(ratio = faults / faults_baseline)
if (nrow(ratios) == 0) {
stop("No ratios computed (missing baseline or zero values)")
}
# Pivot to get aot and interp ratios side by side
ratio_wide <- ratios |>
select(base_name, benchmark, resulttype, variant, ratio) |>
pivot_wider(names_from = variant, values_from = ratio) |>
filter(!is.na(aot), !is.na(interp))
if (nrow(ratio_wide) == 0) {
stop("No paired aot/interp ratios found")
}
# Compute correlation
cor_result <- cor(ratio_wide$aot, ratio_wide$interp, method = "pearson")
cat(sprintf("Pearson correlation: %.4f\n", cor_result))
# Create plot
plot <- ggplot(
ratio_wide,
aes(x = aot, y = interp, color = base_name, shape = resulttype)
) +
geom_point(size = 3, alpha = 0.7) +
scale_x_log10(name = "AOT / C Ratio") +
scale_y_log10(name = "Interpreter / C Ratio") +
labs(
title = sprintf("Ratio Correlation (r = %.4f)", cor_result),
color = "Experiment",
shape = "Marker"
) +
theme_minimal() +
theme(
legend.position = "right",
plot.title = element_text(size = 14, face = "bold")
)
ggsave(
"injections/ratio_correlation_customized.svg",
plot = plot,
width = 10,
height = 8
)
print("Saved ratio_correlation_customized.svg")