Day 7 - Exploratory Data Analysis (EDA)

Fall 2022

Dr. Jared Joseph

September 21, 2022

Overview

Timeline

What is EDA
Numeric Summaries
Visualizing Data
Code Along

Goal

Learn fundamental tools to understand datasets within R.

What is EDA

Common EDA Questions

Data Source

Where is this data from?
Who made it?
For what purpose?
What is the population?
What is the sample like?
What do the values mean?
What is the margin of error?

Data Structure

What data type are each of the variables?
Does anything need re-coded?
What is the missingness like?
What cleaning needs to be done?
Are there outliers?
What do those outliers mean?

Numeric Summaries

Single Number Summaries

Mean: Arithmetic average of a vector, swayed by outliers
Standard Deviation: How spread out the values are in a vector from the mean
Median: Middle number in a vector, but relies on data structure
Mode: Most common value in vector

Multi-Value Summaries

Summary

```{r}
summary(survey$mint_choc)
```

   Min. 1st Qu.  Median    Mean 3rd Qu.    Max. 
    1.0     2.0     4.0     3.6     5.0     5.0

Table

```{r}
table("mint_choc" = survey$mint_choc)
```

mint_choc
1 3 4 5 
4 1 3 7

```{r}
table("mint_choc" = survey$mint_choc,
      "hotdog" = survey$hotdog)
```

         hotdog
mint_choc FALSE TRUE
        1     1    3
        3     1    0
        4     2    1
        5     4    3

Skimr

```{r}
skimr::skim(survey)
```

Data summary
Name	survey
Number of rows	15
Number of columns	23
_______________________
Column type frequency:
character	14
logical	6
numeric	3
________________________
Group variables	None

Variable type: character

skim_variable	n_missing	complete_rate	min	max	n_unique
fav_char	0	1.00	4	39	15
fav_color	1	0.93	3	50	10
b_month	0	1.00	3	9	9
pets	0	1.00	3	17	8
fav_art	3	0.80	17	1158	12
coffee_days	7	0.53	6	62	7
tea_days	10	0.33	6	17	5
soda.pop_days	9	0.40	7	62	5
juice_days	4	0.73	6	62	8
none_days	10	0.33	17	45	5
lt_location	3	0.80	7	47	12
fict	0	1.00	7	11	2
recreation	0	1.00	51	232	15
key	0	1.00	6	41	15

Variable type: logical

skim_variable	n_missing	complete_rate	mean	count
major	15	0	NaN	:
other_classes	15	0	NaN	:
car	0	1	0.33	FAL: 10, TRU: 5
pineapple_pizza	0	1	0.67	TRU: 10, FAL: 5
nerd	0	1	0.73	TRU: 11, FAL: 4
hotdog	0	1	0.47	FAL: 8, TRU: 7

Variable type: numeric

skim_variable	complete_rate	mean	sd	p0	p25	p50	p75	p100	hist
fav_num	1	15.60	14.92	3	7	9	20.5	56	▇▁▂▁▁
mint_choc	1	3.60	1.72	1	2	4	5.0	5	▅▁▁▃▇
hours_sleep	1	6.87	0.74	6	6	7	7.0	8	▆▁▇▁▃

Visualizing Data

Why Visualize?

Bar Chart

```{r}
#| fig-cap: Barplot of Mint x Chocolate Ratings

barplot(table(survey$mint_choc))
```

Barplot of Mint x Chocolate Ratings

Histogram

```{r}
#| fig-cap: Histogram of typical hours sleep

hist(survey$hours_sleep)
```

Histogram of typical hours sleep

Scatterplot

```{r}
#| fig-cap: Scatterplot of Weight x Miles per Gallon

plot(mtcars$wt, mtcars$mpg)
```

Scatterplot of Weight x Miles per Gallon

Line Chart

Boxplot

```{r}
summary(mtcars$mpg)
```

   Min. 1st Qu.  Median    Mean 3rd Qu.    Max. 
  10.40   15.43   19.20   20.09   22.80   33.90

```{r}
#| fig-cap: Boxplot of Miles per gallon

boxplot(mtcars$mpg)
```