MA213 Basic Statistics and Probability - Lab3 Guide

Lab 3: Plotting and Summaries

---

Learning Objectives

• Classify and Analyze Variables: Categorize variables based on their types (e.g., numerical/categorical, continuous/discrete, ordinal), assess their association (positive, negative, or independent), and determine which make sense as explanatory vs. response variables.
• Use R for Data Management and Exploration: Utilize R to load, pre-process, and explore data through visualization and summarization techniques.

---

Graphics with ggplot2 package

Plotting with ggplot2 package begins with

ggplot(data = df, aes(x=x_xvariable, y=y_variable))

where

df : your dataframe name,

aes() : Aesthetics to define -> specifying which variables are mapped to the x and y axes.

and then you add geoms functions – geometrical objects as a graphical representation of the data in the plot (points, lines, bars). ggplot2 offers many different geoms; We will use a few common ones today, including:

• geom_point() : scatter plots, dot plots, etc.

• geom_line() : trend lines, time-series, etc.

• geom_histogram() : histograms

In short,

Plots to show New York Air Quality Data

We want to examine the shape of Ozone to see whether the data is symmetric, skewed or how the mean is centered and where…? Let’s see how this Ozone data is shaped by using several different methods.

Import the Data

df = read.csv("Lab3/airquality.csv")

1. Summary of Ozone.

summary(df$Ozone)

##    Min. 1st Qu.  Median    Mean 3rd Qu.    Max.    NA's 
##    1.00   18.00   31.50   42.13   63.25  168.00      37

2. Stacked dot plot of Ozone .

ggplot(data = df, aes(Ozone)) + 
  geom_dotplot(binwidth = 5) # bins of width 5

## Warning: Removed 37 rows containing missing values or values
## outside the scale range (`stat_bindot()`).

3. Box plot of Ozone.

ggplot(data=df, aes(Ozone)) +
  geom_boxplot()

## Warning: Removed 37 rows containing non-finite outside the
## scale range (`stat_boxplot()`).

This shows that there are two suspected outliers.

4. Histogram of Ozone .

# ggplot version

ggplot(data=df ) + 
  geom_histogram(mapping=aes(Ozone), binwidth = 5) 

## Warning: Removed 37 rows containing non-finite outside the
## scale range (`stat_bin()`).

5. Histogram with different bin option

by choosing how many observations in each bin (e.g. bins=5 means each bin contains 5 observations).

# bins = 5
ggplot(data=df ) + geom_histogram(aes(Ozone), bins = 5)

## Warning: Removed 37 rows containing non-finite outside the
## scale range (`stat_bin()`).

# bins = 30
ggplot(data=df, aes(Ozone)) + 
  geom_histogram(bins = 30) 

## Warning: Removed 37 rows containing non-finite outside the
## scale range (`stat_bin()`).

6. Labels in ggplot

With labs(x='x lable', y='y lable', title = 'name for your plot') you can name your plot.

ggplot(data=df, aes(Ozone)) + 
  geom_histogram(bins = 30, color='yellow') +
  labs(x='Ozone Level', y='Count', title = 'Histogram with 30 bins')

## Warning: Removed 37 rows containing non-finite outside the
## scale range (`stat_bin()`).

7. Scatter plot of Ozone and Temp .

We want to examine whether there is a relationship between ozone and temperature (both numerical). How do we want to approach this?

ggplot(df) + geom_point(aes(x=Ozone, y=Temp)) + 
  labs(title='Scatter plot of Ozone vs Temperature')

## Warning: Removed 37 rows containing missing values or values
## outside the scale range (`geom_point()`).

Titanic Data

We are going to work with titanic data titanic.csv. Let’s import the data.

# loading from csv file
df <- read.csv("Lab3/titanic.csv")
glimpse(df)

## Rows: 2,201
## Columns: 5
## $ X        <int> 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18…
## $ Class    <chr> "3rd", "3rd", "3rd", "3rd", "3rd", "3rd", "3rd", "3rd", "3rd"…
## $ Sex      <chr> "Male", "Male", "Male", "Male", "Male", "Male", "Male", "Male…
## $ Age      <chr> "Child", "Child", "Child", "Child", "Child", "Child", "Child"…
## $ Survived <chr> "No", "No", "No", "No", "No", "No", "No", "No", "No", "No", "…

8. Barplot for Survived

new_df <- df %>% count(Survived)
ggplot(new_df, aes(x = Survived, y=n)) +
  geom_col() 

9. Barplot of Survived and Class

This shows summary table of Survived and Class

df %>% 
  group_by(Survived, Class) %>%
  summarise(count=n()) 

## `summarise()` has grouped output by 'Survived'. You
## can override using the `.groups` argument.

## # A tibble: 8 × 3
## # Groups:   Survived [2]
##   Survived Class count
##   <chr>    <chr> <int>
## 1 No       1st     122
## 2 No       2nd     167
## 3 No       3rd     528
## 4 No       Crew    673
## 5 Yes      1st     203
## 6 Yes      2nd     118
## 7 Yes      3rd     178
## 8 Yes      Crew    212

Adding ggplot() will turn it into barplot using geom_col()

df %>% 
  group_by(Survived, Class) %>%
  summarise(count=n()) %>%
  ggplot(aes(x = factor(Survived), y=count, fill=factor(Class))) +
  geom_col(position = "fill") +  # position can be changed to "dodge", "stack", "jitter", "fill"
  labs(x="Survived")

## `summarise()` has grouped output by 'Survived'. You
## can override using the `.groups` argument.

# for positioning in geom functions
# https://ggplot2.tidyverse.org/reference/layer_positions.html

10. Pie chart of Class

We need table class data first in order to show a pie chart

# dplyr + ggplot

table_data <- df %>% 
  count(Class) %>%
  mutate(Class = factor(Class))

table_data

##   Class   n
## 1   1st 325
## 2   2nd 285
## 3   3rd 706
## 4  Crew 885

# alternate way using group_by() + summarise()

table_data2 <- df %>% 
  group_by(Class) %>%
  summarise(N = n())

table_data2

## # A tibble: 4 × 2
##   Class     N
##   <chr> <int>
## 1 1st     325
## 2 2nd     285
## 3 3rd     706
## 4 Crew    885

Using this table class object table_data, we obtain a pie chart.

ggplot(data = table_data, aes(x="", y=n, fill=Class)) +
  geom_col() +
  geom_text(aes(label = n), position = position_stack(vjust = 0.5)) + 
  coord_polar(theta="y") +
  theme_void()

# or simply 
ggplot(data = table_data, aes(x="", y=n, fill=Class)) +
  geom_col() + 
  coord_polar(theta="y") +
  theme_void()

# using table_data2
ggplot(data = table_data2, aes(x="", y=N, fill=Class)) +
  geom_col() + 
  coord_polar(theta="y") +
  theme_void()