Add File untangling to Figure 3, Table 3, and Table 4

analysis/paper/compare_models.R

@@ -25,11 +25,10 @@ outputFile = args[2]
 # We use R to handle the linear mixed models because Python doesn't support lmms with 2 random effects (cross effects).
 #
 data <- read.csv(inputFile, header = FALSE, col.names = c('Project', 'BugID', 'SmartCommit', 'Flexeme', 'FileUntangling'))
-data <- subset(data, select = -c(FileUntangling))
 data$BugID <- as_factor(data$BugID)
 
 # Convert to long format
-data_long = pivot_longer(data, cols = 3:4, names_to = 'Tool', values_to = 'Performance')
+data_long = pivot_longer(data, cols = all_of(c("SmartCommit", "Flexeme", "FileUntangling")), names_to = 'Tool', values_to = 'Performance')
 
 # Random model
 model_mixed <- lmer(Performance ~ Tool + (1|Project) + (1|BugID), data=data_long)
@@ -50,10 +49,10 @@ rsq(model_simple, adj=TRUE) # Adjusted R^2
 
 # Create Coefficients dataframe
 coeffs = rbind(summary(model_mixed)$coefficients[, "Estimate"], summary(model_simple)$coefficients[, "Estimate"])
-colnames(coeffs) <- c('Flexeme', 'SmartCommit')
+colnames(coeffs) <- c('Flexeme', 'SmartCommit', 'FileUntangling')
 # Create p-values dataframe
 p_vals = rbind(summary(model_mixed)$coefficients[, "Pr(>|t|)"], summary(model_simple)$coefficients[, "Pr(>|t|)"])
-colnames(p_vals) <- c('Flexeme', 'SmartCommit')
+colnames(p_vals) <- c('Flexeme', 'SmartCommit', 'FileUntangling')
 
 tool_stats <- cbind(coeffs, p_vals)
 tool_stats <- rbind(colnames(tool_stats),tool_stats)

analysis/paper/performance_distribution.R

@@ -16,12 +16,11 @@ inputFile = args[1]
 outputFile = args[2]
 
 
-data <- read.csv(inputFile, header = FALSE, col.names = c('Project', 'BugID', 'SmartCommit', 'Flexeme', 'FileUntangling'))
-data <- subset(data, select = -c(FileUntangling))
-data$BugID <- as_factor(data$BugID)
+data <- read.csv(inputFile, header = FALSE, col.names = c('Project', 'Commit', 'SmartCommit', 'Flexeme', 'File_Untangling'))
+data$Commit <- as_factor(data$Commit)
 
 # Convert to long format
-data_long = pivot_longer(data, cols = 3:4, names_to = 'Tool', values_to = 'Performance')
+data_long = pivot_longer(data, cols = c("SmartCommit", "Flexeme", "File_Untangling"), names_to = 'Tool', values_to = 'Performance')
 
 pdf(outputFile)
 flexplot(Performance ~ Tool, data = data_long)

analysis/paper/statistical_analysis_commit_metrics.R

@@ -95,20 +95,26 @@ metrics.data <- read.csv(metrics.path)
 metrics.names <- colnames(metrics.data %>% select(-c('project', 'vid')))
 
 # Join performance with metrics
-performance.metrics <- left_join(performance.data, metrics.data, by = c('project' = 'project', 'bug_id' = 'vid')) %>% select(-c('file_untangling'))
+performance.metrics <- left_join(performance.data, metrics.data, by = c('project' = 'project', 'bug_id' = 'vid'))
 
 # Global
-performance.metrics.long = pivot_longer(performance.metrics, cols = c('smartcommit_rand_index', 'flexeme_rand_index'), names_to = 'Tool', values_to = 'performance')
+performance.metrics.long = pivot_longer(performance.metrics, cols = c('smartcommit_rand_index', 'flexeme_rand_index', 'file_untangling'), names_to = 'Tool', values_to = 'performance')
 summarise_model_all_variables(performance.metrics.long, outputPath, "impact_metrics_all.txt")
 
 # SmartCommit All
-performance.metrics.smartcommit <- select(performance.metrics, -c('flexeme_rand_index'))  %>% rename(performance = smartcommit_rand_index)
+performance.metrics.smartcommit <- select(performance.metrics, -c('flexeme_rand_index', 'file_untangling'))  %>% rename(performance = smartcommit_rand_index)
 summarise_model_all_variables(performance.metrics.smartcommit, outputPath, "impact_metrics_smartcommit_all.txt")
 
 # Flexeme All
-performance.metrics.flexeme <- select(performance.metrics, -c('smartcommit_rand_index'))  %>% rename(performance = flexeme_rand_index)
+performance.metrics.flexeme <- select(performance.metrics, -c('smartcommit_rand_index', 'file_untangling'))  %>% rename(performance = flexeme_rand_index)
 summarise_model_all_variables(performance.metrics.flexeme, outputPath, "impact_metrics_flexeme_all.txt")
 
+# File Untangling ALL
+performance.metrics.file <- select(performance.metrics, -c('smartcommit_rand_index', 'flexeme_rand_index'))  %>% rename(performance = file_untangling)
+summarise_model_all_variables(performance.metrics.file, outputPath, "impact_metrics_file_all.txt")
+
 # For each metric, do model analysis and print graph in pdf
 generate_pairwise_analysis(performance.metrics.smartcommit, metrics.names, "smartcommit", outputPath)
 generate_pairwise_analysis(performance.metrics.flexeme, metrics.names, "flexeme", outputPath)
+generate_pairwise_analysis(performance.metrics.file.smartcommit, metrics.names, "file-smartcommit", outputPath)
+generate_pairwise_analysis(performance.metrics.file.flexeme, metrics.names, "file-flexeme", outputPath)