# sets the directory of location of this script as the current directory
# setwd(dirname(rstudioapi::getSourceEditorContext()$path))
### load packages
require(pacman)
p_load('jsonlite', 'xlsx', 'ggpubr',
'stargazer', 'DT', 'tools', 'tidyverse')
### load abbrevations dictionary
setwd("data")
abbreviations_dict <- jsonlite::fromJSON(txt = "abbreviations_dict.txt")
### load frequency table of codes
setwd("../../output/G3")
#> RR
codes_frequency_RR <- xlsx::read.xlsx2(file = "rescue robot frequency table codes.xlsx", sheetIndex = 1)
codes_frequency_RR$Code <- str_replace_all(
string = codes_frequency_RR$Code,
pattern = "_",
replacement = " ")
codes_frequency_RR$Code <- tools::toTitleCase(codes_frequency_RR$Code)
codes_frequency_RR$soft <- as.numeric(codes_frequency_RR$soft)
codes_frequency_RR$rigid <- as.numeric(codes_frequency_RR$rigid)
codes_frequency_RR$soft <- codes_frequency_RR$soft + codes_frequency_RR$rigid
#> SAR
codes_frequency_SAR <- xlsx::read.xlsx2(file = "socially assistive robot frequency table codes.xlsx", sheetIndex = 1)
codes_frequency_SAR$Code <- str_replace_all(
string = codes_frequency_SAR$Code,
pattern = "_",
replacement = " ")
codes_frequency_SAR$Code <- tools::toTitleCase(codes_frequency_SAR$Code)
codes_frequency_SAR$soft <- as.numeric(codes_frequency_SAR$soft)
codes_frequency_SAR$rigid <- as.numeric(codes_frequency_SAR$rigid)
codes_frequency_SAR$soft <- codes_frequency_SAR$soft + codes_frequency_SAR$rigid
### load improved coding guidelines
setwd("json RR - improved")
# List all JSON files in the directory
json_files <- list.files(pattern = "json$", full.names = FALSE)
# Initialize an empty list to store code descriptions
list_json_files_RR <- list()
# Loop through each JSON file
for (i in 1:length(json_files)) {
# Read the JSON file
list_json_files_RR[[i]] <- jsonlite::fromJSON(json_files[i])
}
names(list_json_files_RR) <- json_files
setwd("../json SAR - improved")
# List all JSON files in the directory
json_files <- list.files(pattern = "json$", full.names = FALSE)
# Initialize an empty list to store code descriptions
list_json_files_SAR <- list()
# Loop through each JSON file
for (i in 1:length(json_files)) {
# Read the JSON file
list_json_files_SAR[[i]] <- jsonlite::fromJSON(json_files[i])
}
names(list_json_files_SAR) <- json_files
rm(i); rm(json_files)Create tables and graphics for G3 (result part II)
Notes
load cleaned data files
Create table for Supplementary Materials
for RR
# Create an empty data frame to store results
results_df <- data.frame(
CategoryAbbrevation = character(),
Category = character(),
CodeName = character(),
Description = character(),
stringsAsFactors = FALSE
)
# Loop through the JSON files
for (i in 1:length(list_json_files_RR)) {
# Improve code names
names(list_json_files_RR[[i]]$code_descriptions) <- str_replace_all(
string = names(list_json_files_RR[[i]]$code_descriptions),
pattern = "_",
replacement = " "
)
names(list_json_files_RR[[i]]$code_descriptions) <- tools::toTitleCase(
names(list_json_files_RR[[i]]$code_descriptions)
)
# Get full name of category
tmp_categoryAbbrevation <- names(abbreviations_dict$abbreviations_dict[
names(abbreviations_dict$abbreviations_dict) == list_json_files_RR[[i]]$category])
tmp_category <- abbreviations_dict$abbreviations_dict[
names(abbreviations_dict$abbreviations_dict) == list_json_files_RR[[i]]$category
][[1]]
# Loop through each code description
for (j in 1:length(list_json_files_RR[[i]]$code_descriptions)) {
# Extract the code name and description
code_name <- names(list_json_files_RR[[i]]$code_descriptions)[j]
description <- list_json_files_RR[[i]]$code_descriptions[j][[1]]
# Append to the results data frame
results_df <- rbind(
results_df,
data.frame(
CategoryAbbrevation = tmp_categoryAbbrevation,
Category = tmp_category,
CodeName = code_name,
Description = description,
stringsAsFactors = FALSE
)
)
}
}
colnames(results_df) <- c("Abbrevation", "Category", "Code", "Description")
results_df$soft <- NA
results_df$rigid <- NA
### add frequencies
for(c in unique(results_df$Abbrevation)){
tmp_codes_frequency <- codes_frequency_RR[codes_frequency_RR$Category == c, ]
for(i in 1:nrow(tmp_codes_frequency)){
results_df$soft[results_df$Abbrevation == c & results_df$Code == tmp_codes_frequency[i,"Code"]] <- tmp_codes_frequency[i,"soft"]
results_df$rigid[results_df$Abbrevation == c & results_df$Code == tmp_codes_frequency[i,"Code"]] <- tmp_codes_frequency[i,"rigid"]
}
}
### save overall table
xlsx::write.xlsx2(x = results_df, file = "outputs/RR_codes_complete.xlsx")
### sort according to frequency soft, and save top 20
results_df_sorted <- results_df[order(results_df$soft, decreasing = TRUE),]
xlsx::write.xlsx2(x = results_df_sorted[1:20, ], file = "outputs/RR_codes_first20.xlsx")
### backup:
results_df_RR <- results_df
### show overall table
DT::datatable(data = results_df_RR)for SAR
# Create an empty data frame to store results
results_df <- data.frame(
CategoryAbbrevation = character(),
Category = character(),
CodeName = character(),
Description = character(),
stringsAsFactors = FALSE
)
# Loop through the JSON files
for (i in 1:length(list_json_files_SAR)) {
# Improve code names
names(list_json_files_SAR[[i]]$code_descriptions) <- str_replace_all(
string = names(list_json_files_SAR[[i]]$code_descriptions),
pattern = "_",
replacement = " "
)
names(list_json_files_SAR[[i]]$code_descriptions) <- tools::toTitleCase(
names(list_json_files_SAR[[i]]$code_descriptions)
)
# Get full name of category
tmp_categoryAbbrevation <- names(abbreviations_dict$abbreviations_dict[
names(abbreviations_dict$abbreviations_dict) == list_json_files_SAR[[i]]$category])
tmp_category <- abbreviations_dict$abbreviations_dict[
names(abbreviations_dict$abbreviations_dict) == list_json_files_SAR[[i]]$category
][[1]]
# Loop through each code description
for (j in 1:length(list_json_files_SAR[[i]]$code_descriptions)) {
# Extract the code name and description
code_name <- names(list_json_files_SAR[[i]]$code_descriptions)[j]
description <- list_json_files_SAR[[i]]$code_descriptions[j][[1]]
# Append to the results data frame
results_df <- rbind(
results_df,
data.frame(
CategoryAbbrevation = tmp_categoryAbbrevation,
Category = tmp_category,
CodeName = code_name,
Description = description,
stringsAsFactors = FALSE
)
)
}
}
colnames(results_df) <- c("Abbrevation", "Category", "Code", "Description")
results_df$soft <- NA
results_df$rigid <- NA
### add frequencies
for(c in unique(results_df$Abbrevation)){
tmp_codes_frequency <- codes_frequency_SAR[codes_frequency_SAR$Category == c, ]
for(i in 1:nrow(tmp_codes_frequency)){
results_df$soft[results_df$Abbrevation == c & results_df$Code == tmp_codes_frequency[i,"Code"]] <- tmp_codes_frequency[i,"soft"]
results_df$rigid[results_df$Abbrevation == c & results_df$Code == tmp_codes_frequency[i,"Code"]] <- tmp_codes_frequency[i,"rigid"]
}
}
### save overall table
xlsx::write.xlsx2(x = results_df, file = "outputs/SAR_codes_complete.xlsx")
### sort according to frequency soft, and save top 20
results_df_sorted <- results_df[order(results_df$soft, decreasing = TRUE),]
xlsx::write.xlsx2(x = results_df_sorted[1:20, ], file = "outputs/SAR_codes_first20.xlsx")
### backup:
results_df_SAR <- results_df
### show overall table
DT::datatable(data = results_df_SAR)Create plots for Main Article
for RR
filter for most frequent soft terms:
for(c in unique(results_df_RR$Abbrevation)){
tmp_frequency <- results_df_RR[results_df_RR$Abbrevation == c, ]
tmp_frequency <- tmp_frequency[!is.na(tmp_frequency$soft),]
tmp_frequency <- tmp_frequency[tmp_frequency$soft >= median(tmp_frequency$soft, na.rm = TRUE),]
tmp_frequency$Abbrevation <- NULL
tmp_frequency$Description <- NULL
tmp_title <- paste0(unique(tmp_frequency$Category), collapse = " ") # " - rescue robots"
cat("Plot for:\n", tmp_title, "\n")
# Insert line breaks into long category names
tmp_frequency$Code <- str_wrap(tmp_frequency$Code, width = 20)
# Reshape data for ggplot (long format)
tmp_long <- reshape2::melt(tmp_frequency, id.vars = c("Category", "Code"),
variable.name = "Type", value.name = "Count")
# Create the plot
plot <- ggplot(tmp_long, aes(y = factor(Code, levels = rev(unique(Code))),
x = Count, fill = Type)) +
geom_bar(stat = "identity", position = position_dodge(width = 0.7)) + # Increase bar spacing
labs(
title = tmp_title,
y = "Category",
x = "Frequency",
fill = "Type"
) +
theme_pubr() + # APA-like theme
theme(
plot.title = element_text(size = 16, face = "bold"),
axis.title = element_text(size = 14),
axis.title.y = element_blank(),
axis.text.y = element_text(size = 12, hjust = 1), # Adjust y-axis text alignment
axis.text.x = element_text(size = 14),
legend.title = element_text(size = 12),
legend.text = element_text(size = 10),
panel.grid.major = element_blank(), # Remove grid lines for clarity
panel.grid.minor = element_blank()
) +
scale_y_discrete(labels = function(labels) str_wrap(labels, width = 20)) # Ensure proper wrapping
# Display the plot
print(plot)
}Plot for:
perceived negative anthropomorphism 
Plot for:
perceived positive anthropomorphism 
Plot for:
perceived negative Human-Robot-Interaction 
Plot for:
perceived positive Human-Robot-Interaction 
Plot for:
perceived risks 
Plot for:
perceived safety 
Plot for:
perceived technological limitations 
Plot for:
perceived technological possibilities 
do not filter for most frequent soft terms:
setwd("outputs/G3_RR")
for(c in unique(results_df_RR$Abbrevation)){
tmp_frequency <- results_df_RR[results_df_RR$Abbrevation == c, ]
tmp_frequency <- tmp_frequency[!is.na(tmp_frequency$soft),]
# tmp_frequency <- tmp_frequency[tmp_frequency$soft >= median(tmp_frequency$soft, na.rm = TRUE),]
tmp_frequency$Abbrevation <- NULL
tmp_frequency$Description <- NULL
tmp_title <- paste0(unique(tmp_frequency$Category), collapse = " ") # " - rescue robots"
cat("Plot for:\n", tmp_title, "\n")
# Insert line breaks into long category names
tmp_frequency$Code <- str_wrap(tmp_frequency$Code, width = 20)
# Reshape data for ggplot (long format)
tmp_long <- reshape2::melt(tmp_frequency, id.vars = c("Category", "Code"),
variable.name = "Type", value.name = "Count")
# Create the plot
plot <- ggplot(tmp_long, aes(y = factor(Code, levels = rev(unique(Code))),
x = Count, fill = Type)) +
geom_bar(stat = "identity", position = position_dodge(width = 0.7)) + # Increase bar spacing
labs(
title = tmp_title,
y = "Category",
x = "Frequency",
fill = "Type"
) +
theme_pubr() + # APA-like theme
theme(
plot.title = element_text(size = 16, face = "bold"),
axis.title = element_text(size = 14),
axis.title.y = element_blank(),
axis.text.y = element_text(size = 12, hjust = 1), # Adjust y-axis text alignment
axis.text.x = element_text(size = 14),
legend.title = element_text(size = 12),
legend.text = element_text(size = 10),
panel.grid.major = element_blank(), # Remove grid lines for clarity
panel.grid.minor = element_blank()
) +
scale_y_discrete(labels = function(labels) str_wrap(labels, width = 20)) # Ensure proper wrapping
# Display the plot
print(plot)
ggsave(paste0(tmp_title, ".pdf"))
}Plot for:
perceived negative anthropomorphism 
Saving 7 x 5 in imagePlot for:
perceived positive anthropomorphism
Saving 7 x 5 in imagePlot for:
perceived negative Human-Robot-Interaction 
Saving 7 x 5 in imagePlot for:
perceived positive Human-Robot-Interaction 
Saving 7 x 5 in imagePlot for:
perceived risks 
Saving 7 x 5 in imagePlot for:
perceived safety 
Saving 7 x 5 in imagePlot for:
perceived technological limitations 
Saving 7 x 5 in imagePlot for:
perceived technological possibilities 
Saving 7 x 5 in imagefor SAR
filter for most frequent soft terms:
for(c in unique(results_df_SAR$Abbrevation)){
tmp_frequency <- results_df_SAR[results_df_SAR$Abbrevation == c, ]
tmp_frequency <- tmp_frequency[!is.na(tmp_frequency$soft),]
tmp_frequency <- tmp_frequency[tmp_frequency$soft >= median(tmp_frequency$soft, na.rm = TRUE),]
tmp_frequency$Abbrevation <- NULL
tmp_frequency$Description <- NULL
tmp_title <- paste0(unique(tmp_frequency$Category), collapse = " ") # " - rescue robots"
cat("Plot for:\n", tmp_title, "\n")
# Insert line breaks into long category names
tmp_frequency$Code <- str_wrap(tmp_frequency$Code, width = 20)
# Reshape data for ggplot (long format)
tmp_long <- reshape2::melt(tmp_frequency, id.vars = c("Category", "Code"),
variable.name = "Type", value.name = "Count")
# Create the plot
plot <- ggplot(tmp_long, aes(y = factor(Code, levels = rev(unique(Code))),
x = Count, fill = Type)) +
geom_bar(stat = "identity", position = position_dodge(width = 0.7)) + # Increase bar spacing
labs(
title = tmp_title,
y = "Category",
x = "Frequency",
fill = "Type"
) +
theme_pubr() + # APA-like theme
theme(
plot.title = element_text(size = 16, face = "bold"),
axis.title = element_text(size = 14),
axis.title.y = element_blank(),
axis.text.y = element_text(size = 12, hjust = 1), # Adjust y-axis text alignment
axis.text.x = element_text(size = 14),
legend.title = element_text(size = 12),
legend.text = element_text(size = 10),
panel.grid.major = element_blank(), # Remove grid lines for clarity
panel.grid.minor = element_blank()
) +
scale_y_discrete(labels = function(labels) str_wrap(labels, width = 20)) # Ensure proper wrapping
# Display the plot
print(plot)
}Plot for:
perceived negative anthropomorphism 
Plot for:
perceived positive anthropomorphism 
Plot for:
perceived negative Human-Robot-Interaction 
Plot for:
perceived positive Human-Robot-Interaction 
Plot for:
perceived risks 
Plot for:
perceived safety 
Plot for:
perceived technological limitations 
Plot for:
perceived technological possibilities 
do not filter for most frequent soft terms:
setwd("outputs/G3_SAR")
for(c in unique(results_df_SAR$Abbrevation)){
tmp_frequency <- results_df_SAR[results_df_SAR$Abbrevation == c, ]
tmp_frequency <- tmp_frequency[!is.na(tmp_frequency$soft),]
# tmp_frequency <- tmp_frequency[tmp_frequency$soft >= median(tmp_frequency$soft, na.rm = TRUE),]
tmp_frequency$Abbrevation <- NULL
tmp_frequency$Description <- NULL
tmp_title <- paste0(unique(tmp_frequency$Category), collapse = " ") # " - rescue robots"
cat("Plot for:\n", tmp_title, "\n")
# Insert line breaks into long category names
tmp_frequency$Code <- str_wrap(tmp_frequency$Code, width = 20)
# Reshape data for ggplot (long format)
tmp_long <- reshape2::melt(tmp_frequency, id.vars = c("Category", "Code"),
variable.name = "Type", value.name = "Count")
# Create the plot
plot <- ggplot(tmp_long, aes(y = factor(Code, levels = rev(unique(Code))),
x = Count, fill = Type)) +
geom_bar(stat = "identity", position = position_dodge(width = 0.7)) + # Increase bar spacing
labs(
title = tmp_title,
y = "Category",
x = "Frequency",
fill = "Type"
) +
theme_pubr() + # APA-like theme
theme(
plot.title = element_text(size = 16, face = "bold"),
axis.title = element_text(size = 14),
axis.title.y = element_blank(),
axis.text.y = element_text(size = 12, hjust = 1), # Adjust y-axis text alignment
axis.text.x = element_text(size = 14),
legend.title = element_text(size = 12),
legend.text = element_text(size = 10),
panel.grid.major = element_blank(), # Remove grid lines for clarity
panel.grid.minor = element_blank()
) +
scale_y_discrete(labels = function(labels) str_wrap(labels, width = 20)) # Ensure proper wrapping
# Display the plot
print(plot)
ggsave(paste0(tmp_title, ".pdf"))
}Plot for:
perceived negative anthropomorphism 
Saving 7 x 5 in imagePlot for:
perceived positive anthropomorphism 
Saving 7 x 5 in imagePlot for:
perceived negative Human-Robot-Interaction 
Saving 7 x 5 in imagePlot for:
perceived positive Human-Robot-Interaction 
Saving 7 x 5 in imagePlot for:
perceived risks 
Saving 7 x 5 in imagePlot for:
perceived safety 
Saving 7 x 5 in imagePlot for:
perceived technological limitations 
Saving 7 x 5 in imagePlot for:
perceived technological possibilities 
Saving 7 x 5 in image