Introduction

This project analyzes Foodpanda order data using R, divided into key sections: basic understanding, aggregation, filtering, trends, business insights, pie chart analysis, box plot analysis, linear regression, and correlation analysis. Using dplyr, ggplot2, and reshape2, the data is processed, analyzed, and visualized to understand order patterns, customer ratings, revenue distribution, and relationships between variables.

Setup & Data Engineering

library(dplyr)
## 
## Attaching package: 'dplyr'
## The following objects are masked from 'package:stats':
## 
##     filter, lag
## The following objects are masked from 'package:base':
## 
##     intersect, setdiff, setequal, union
library(ggplot2)
library(reshape2)

df <- read.csv("Foodpanda Analysis Dataset.csv")

# Creates order_value = price × quantity
df <- df %>% mutate(order_value = price * quantity)

1. BASIC UNDERSTANDING

1.1 What is the total number of orders?

nrow(df)
## [1] 6000
# Counts total number of rows → total orders

Conclusion: Shows the overall scale of the dataset and business activity.

1.2 How many unique restaurants are on the platform?

df %>% summarise(unique_restaurants = n_distinct(restaurant_name))
##   unique_restaurants
## 1                  5
# Counts distinct restaurants

Conclusion: Indicates platform diversity and competition level.

1.4 What is the average customer rating?

df %>% summarise(avg_rating = mean(rating, na.rm = TRUE))
##   avg_rating
## 1   2.996833
# Calculates mean rating

Conclusion: Reflects overall customer satisfaction on the platform.

1.5 Which cities generate the most orders?

city_orders <- df %>% count(city, sort = TRUE)

ggplot(city_orders, aes(x = reorder(city, n), y = n)) +
  geom_bar(stat = "identity") +
  coord_flip() +
  theme_minimal()

# Orders grouped by city

Conclusion: Identifies high-demand cities driving platform usage.

2. AGGREGATION

2.1 Which restaurants have the highest average ratings?

rating_rest <- df %>%
  group_by(restaurant_name) %>%
  summarise(avg_rating = mean(rating, na.rm = TRUE)) %>%
  arrange(desc(avg_rating)) %>%
  head(10)

ggplot(rating_rest, aes(x = reorder(restaurant_name, avg_rating), y = avg_rating)) +
  geom_bar(stat = "identity") +
  coord_flip() +
  theme_minimal()

# Average rating per restaurant

Conclusion: Highlights top-performing restaurants in terms of quality.

2.2 Which restaurant has the highest average order value?

df %>%
  group_by(restaurant_name) %>%
  summarise(avg_value = mean(order_value, na.rm = TRUE)) %>%
  arrange(desc(avg_value)) %>%
  head(1)
## # A tibble: 1 × 2
##   restaurant_name avg_value
##   <chr>               <dbl>
## 1 Pizza Hut           2453.
# Finds highest spending restaurant

Conclusion: Identifies premium restaurants with higher customer spending.

2.3 Which cities generate the most revenue?

city_revenue <- df %>%
  group_by(city) %>%
  summarise(revenue = sum(order_value, na.rm = TRUE))

ggplot(city_revenue, aes(x = reorder(city, revenue), y = revenue)) +
  geom_bar(stat = "identity") +
  coord_flip() +
  theme_minimal()

# Revenue aggregated by city

Conclusion: Shows which locations contribute most to total revenue.

2.4 Which cuisine categories generate the most revenue?

cat_revenue <- df %>%
  group_by(category) %>%
  summarise(revenue = sum(order_value, na.rm = TRUE))

ggplot(cat_revenue, aes(x = reorder(category, revenue), y = revenue)) +
  geom_bar(stat = "identity") +
  coord_flip() +
  theme_minimal()

# Revenue by category

Conclusion: Reveals which cuisines are most profitable.

6. PROPORTION ANALYSIS

6.1 What is the distribution of customer satisfaction levels?

rating_tiers <- df %>%
  mutate(tier = case_when(
    rating >= 4.5 ~ "Excellent",
    rating >= 3.0 ~ "Average",
    TRUE ~ "Poor"
  )) %>%
  count(tier)

ggplot(rating_tiers, aes(x = "", y = n, fill = tier)) +
  geom_bar(stat = "identity", width = 1) +
  coord_polar("y") +
  theme_void()

# Rating distribution

Conclusion: Shows proportion of good vs poor customer experiences.

6.2 How does revenue split between single and multi-item orders?

order_size_rev <- df %>%
  mutate(size_type = ifelse(quantity == 1, "Single", "Multi")) %>%
  group_by(size_type) %>%
  summarise(revenue = sum(order_value, na.rm = TRUE))

ggplot(order_size_rev, aes(x = "", y = revenue, fill = size_type)) +
  geom_bar(stat = "identity", width = 1) +
  coord_polar("y") +
  theme_void()

# Revenue split by order size

Conclusion: Indicates whether bulk orders contribute more revenue.

7. DISTRIBUTION ANALYSIS

7.1 How does order value vary across cuisine categories?

upper_limit <- quantile(df$order_value, 0.95, na.rm = TRUE)

ggplot(df, aes(x = category, y = order_value)) +
  geom_boxplot() +
  coord_flip(ylim = c(0, upper_limit)) +
  theme_minimal()

# Boxplot of order values

Conclusion: Highlights spending variability and outliers across cuisines.

7.2 How do item prices vary across cuisine categories?

price_upper_limit <- quantile(df$price, 0.95, na.rm = TRUE)

ggplot(df, aes(x = category, y = price)) +
  geom_boxplot() +
  coord_flip(ylim = c(0, price_upper_limit)) +
  theme_minimal()

# Price distribution

Conclusion: Differentiates budget vs premium cuisines.

8. RELATIONSHIP ANALYSIS

8.1 Do expensive restaurants get better ratings?

rest_price_rating <- df %>%
  group_by(restaurant_name) %>%
  summarise(
    avg_item_price = mean(price, na.rm = TRUE),
    avg_rating = mean(rating, na.rm = TRUE),
    total_orders = n()
  ) %>%
  filter(total_orders > 5)

ggplot(rest_price_rating, aes(x = avg_item_price, y = avg_rating)) +
  geom_point() +
  geom_smooth(method = "lm") +
  theme_minimal()
## `geom_smooth()` using formula = 'y ~ x'

# Price vs rating regression

Conclusion: Shows whether higher prices correlate with better ratings.

8.2 Does popularity (orders) indicate better quality (ratings)?

rest_performance <- df %>%
  group_by(restaurant_name) %>%
  summarise(total_orders = n(), avg_rating = mean(rating, na.rm = TRUE))

ggplot(rest_performance, aes(x = total_orders, y = avg_rating)) +
  geom_point() +
  geom_smooth(method = "lm") +
  theme_minimal()
## `geom_smooth()` using formula = 'y ~ x'

# Orders vs rating relationship

Conclusion: Evaluates if popular restaurants maintain higher ratings.

9. CORRELATION ANALYSIS

9.1 How can we create key restaurant performance metrics?

restaurant_kpis <- df %>%
  group_by(restaurant_name) %>%
  summarise(
    Total_Revenue = sum(order_value, na.rm = TRUE),
    Total_Orders = n(),
    Avg_Price = mean(price, na.rm = TRUE),
    Avg_Rating = mean(rating, na.rm = TRUE)
  ) %>%
  select(-restaurant_name)
# KPI creation

Conclusion: Converts raw data into meaningful business metrics.

9.2 How are these business metrics correlated?

kpi_cor_matrix <- cor(restaurant_kpis, use = "complete.obs")
print(kpi_cor_matrix)
##               Total_Revenue Total_Orders  Avg_Price Avg_Rating
## Total_Revenue     1.0000000    0.9505147  0.9123879  0.1572310
## Total_Orders      0.9505147    1.0000000  0.8142577  0.4372608
## Avg_Price         0.9123879    0.8142577  1.0000000 -0.1399778
## Avg_Rating        0.1572310    0.4372608 -0.1399778  1.0000000
# Correlation matrix

Conclusion: Shows relationships between revenue, orders, price, and ratings.

9.3 What insights can we derive from the correlation heatmap?

kpi_melt <- melt(kpi_cor_matrix)

ggplot(kpi_melt, aes(x = Var1, y = Var2, fill = value)) +
  geom_tile() +
  geom_text(aes(label = round(value, 2))) +
  theme_minimal()

# Heatmap visualization

Conclusion: Helps quickly identify strong positive or negative relationships.

Conclusion

The analysis across all sections highlights important patterns in orders, ratings, and revenue. It identifies top-performing categories and cities, detects variability and outliers, and explores relationships between variables using regression and correlation. Overall, the project demonstrates how structured data analysis can generate useful business insights. ````