Upload New File

# clear workspace

rm(list = ls())

#TASK 2

# load necessary packages

library(httr)

library(jsonlite)

library(lubridate)

#Step 1

#get 100 games data- GET() send a request to the API endpoint- you get raw data in JSON format

top100.request <- GET(url = "https://steamspy.com/api.php?request=top100forever")

#Step 2

#we need to read the response as a string. content(..,as="text") extract the body of response as readable text. UTF-8 ensures special characters are handled corrrectly.

top100.raw <- content(top100.request, as = "text", encoding = "UTF-8")

#Step 3

#Transform JSON into an R object.

top100.list <- fromJSON(top100.raw)

#Step 4: Convert list into a data frame

# Can't make data frame directly as list is nested. Therefore, use lapply() that convert each game entry to small data frame.

top100.list1 <- lapply(top100.list, as.data.frame)

#do.call("rbind",..)- stacks them row by row into one big data frame

top100 <- do.call("rbind", top100.list1)

#Step 5 Inspect the data

head(top100) #First 6 rows

summary (top100)

#Step 6 Transform price-related variables

#Prices are in cents and stored as text. We need to convert them to numeric and scale to dollars.

top100$price <- as.numeric(top100$price)*0.01 #multiply 0.01 changes cents to dollars

top100$initialprice <- as.numeric(top100$initialprice)*0.01 #as.numerics converts text to numbers

top100$discount <- as.numeric(top100$discount)

#Step 7

gameid <- top100$appid[1]

game_name <- top100$name[1]

#TASK 3

#API endpoint for game details

details.url <- "http://store.steampowered.com/api/appdetails/?"

#API expects the game ID in a query list

query.list <- list(appids=gameid) #gameid is the variable and wrapping it in list() make it compatible with Get().

#Perform GET request- fetch raw details of the game and response is again in binary.

details.request <-GET(url=details.url, query=query.list)

#We need response in string, therefore converting the raw response into readable text.

details.raw <- content(details.request, as="text", encoding="UTF-8")

#we transform JSON to R list

details.list <- fromJSON(details.raw)

#The useful information is indie the $data element.

details <- details.list[[1]]$data #details.list[[1]]- select first element of the list. $data- extracts the structured information about the game.

details$release_date$date

details$genres

details$required_age

details$short_description

#TASK 4 (Collect user reviews)

#STEP 1

rev.url1 <- "https://store.steampowered.com/appreviews/"

rev.url2 <- "?json=1&filter=recent&num_per_page=100&cursor="

cursor <- "*"

#rev.url1- base url for reviews

#rev.url2- specifies JSON output, filter for recent reviews, limit to 100 reviews, and cursor for pagination.

#cursor- means start at the beginning.

#Step 2 (Perform GET request)

rev.request <- GET(url =paste0(rev.url1, gameid, rev.url2, cursor))

#paste0()- concatenates strings without spaces i.e. buils full URL

#Step 3 (Convert binary content to text)

rev.raw <- content(rev.request, as="text", encoding="UTF-8")

#Step 4 (JSON to R list)

rev.list <- fromJSON(rev.raw)

#Step 5 (Extract reviews into a data frame)

reviews <- rev.list$reviews

#Step 6 (Inspect the data)

names(reviews) #Column names

summary(reviews) #summary statistics- author data is nested

#Step 7 (Unnest the author data)

reviews <- do.call("data.frame", reviews) #expand nested structure into proper columns

names(reviews)

#Step 8 (Fix data types)

# Convert UNIX timestamp to readable date-time

reviews$timestamp_created <- as_datetime(reviews$timestamp_created)

reviews$timestamp_created

#Ensure votes and playtime are numeric

reviews$votes_up <- as.numeric(reviews$votes_up)

reviews$votes_up

reviews$author.num_games_owned <- as.numeric(reviews$author.num_games_owned)

reviews$author.num_games_owned

reviews

reviews <- unique (reviews)

reviews

 No newline at end of file