Part 4: Exploratory Data Analysis & Data Visualization¶

Introduction to Python¶

Tom Paskhalis¶

2022-07-27¶
Data Science Summer School 2022¶

Exploratory data analysis¶

Exploratory data analysis can never be the whole story, but nothing else can serve as the foundation stone - as the first step.

Tukey, 1977

  • Exploratory data analysis (EDA) is the first and often most important step in research
  • Study's feasibility, scope and framing would usually depend on its results
  • Specific details of EDA depend upon the type and quality of data available

Measurement scales¶

Source: Stevens (1946)

Measurement scales in Pandas¶

  • The 4 measurement scales defined by Stevens (1946) can be roughly represented in pandas as follows:
  • Interval and ratio -> numeric
  • Nominal and ordinal -> categorical

Loading the dataset¶

In [1]:
import pandas as pd
In [2]:
united_2022 = pd.read_csv('../data/united_2022.csv')
In [3]:
united_2022.head(n = 1)
Out[3]:
Carrier Code Date (MM/DD/YYYY) Flight Number Tail Number Origin Airport Destination Airport Scheduled departure time Actual departure time Scheduled elapsed time (Minutes) Actual elapsed time (Minutes) Departure delay (Minutes) Wheels-off time Taxi-Out time (Minutes) Delay Carrier (Minutes) Delay Weather (Minutes) Delay National Aviation System (Minutes) Delay Security (Minutes) Delay Late Aircraft Arrival (Minutes)
0 UA 01/01/2022 225 N488UA ATL DEN 16:15:00 17:23:00 211 240 68 17:35:00 12 0 0 29 0 68

Summarizing numeric variables¶

  • DataFrame methods in pandas can automatically handle (exclude) missing data (NaN)
In [4]:
united_2022.describe() # DataFrame.describe() provides an range of summary statistics
Out[4]:
Flight Number Scheduled elapsed time (Minutes) Actual elapsed time (Minutes) Departure delay (Minutes) Taxi-Out time (Minutes) Delay Carrier (Minutes) Delay Weather (Minutes) Delay National Aviation System (Minutes) Delay Security (Minutes) Delay Late Aircraft Arrival (Minutes)
count 154575.000000 154575.000000 154575.000000 154575.000000 154575.000000 154575.000000 154575.000000 154575.000000 154575.000000 154575.000000
mean 1442.155795 192.808876 180.380618 14.024842 18.978451 5.010209 0.552172 2.590923 0.001844 5.616122
std 732.636240 86.048770 88.502329 46.833963 10.081536 25.454662 12.039743 13.370774 0.281220 28.164622
min 42.000000 42.000000 0.000000 -29.000000 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000
25% 691.000000 134.000000 123.000000 -4.000000 14.000000 0.000000 0.000000 0.000000 0.000000 0.000000
50% 1546.000000 173.000000 164.000000 0.000000 17.000000 0.000000 0.000000 0.000000 0.000000 0.000000
75% 2060.000000 239.000000 230.000000 12.000000 21.000000 0.000000 0.000000 0.000000 0.000000 0.000000
max 3010.000000 685.000000 711.000000 1403.000000 184.000000 1215.000000 876.000000 999.000000 70.000000 1280.000000

Methods for summarizing numeric variables¶

In [5]:
# Rather than using describe(), we can apply individual methods
united_2022['Departure delay (Minutes)'].mean()
Out[5]:
14.024842309558467
In [6]:
united_2022['Departure delay (Minutes)'].median() # Median
Out[6]:
0.0
In [7]:
united_2022['Departure delay (Minutes)'].std() # Standard deviation
Out[7]:
46.833962555880554
In [8]:
import statistics ## We don't have to rely only on methods provided by `pandas`
statistics.stdev(united_2022.iloc[:,10])
Out[8]:
46.833962555880554

Summarizing categorical variables¶

In [9]:
united_2022.describe(include = 'all') # Adding include = 'all' tells pandas to summarize all variables
Out[9]:
Carrier Code Date (MM/DD/YYYY) Flight Number Tail Number Origin Airport Destination Airport Scheduled departure time Actual departure time Scheduled elapsed time (Minutes) Actual elapsed time (Minutes) Departure delay (Minutes) Wheels-off time Taxi-Out time (Minutes) Delay Carrier (Minutes) Delay Weather (Minutes) Delay National Aviation System (Minutes) Delay Security (Minutes) Delay Late Aircraft Arrival (Minutes)
count 154575 154575 154575.000000 151157 154575 154575 154575 154575 154575.000000 154575.000000 154575.000000 154575 154575.000000 154575.000000 154575.000000 154575.000000 154575.000000 154575.000000
unique 1 151 NaN 747 12 110 1066 1307 NaN NaN NaN 1305 NaN NaN NaN NaN NaN NaN
top UA 05/26/2022 NaN N32404 DEN EWR 06:00:00 00:00:00 NaN NaN NaN 00:00:00 NaN NaN NaN NaN NaN NaN
freq 154575 1143 NaN 337 32245 9800 1686 3418 NaN NaN NaN 3473 NaN NaN NaN NaN NaN NaN
mean NaN NaN 1442.155795 NaN NaN NaN NaN NaN 192.808876 180.380618 14.024842 NaN 18.978451 5.010209 0.552172 2.590923 0.001844 5.616122
std NaN NaN 732.636240 NaN NaN NaN NaN NaN 86.048770 88.502329 46.833963 NaN 10.081536 25.454662 12.039743 13.370774 0.281220 28.164622
min NaN NaN 42.000000 NaN NaN NaN NaN NaN 42.000000 0.000000 -29.000000 NaN 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000
25% NaN NaN 691.000000 NaN NaN NaN NaN NaN 134.000000 123.000000 -4.000000 NaN 14.000000 0.000000 0.000000 0.000000 0.000000 0.000000
50% NaN NaN 1546.000000 NaN NaN NaN NaN NaN 173.000000 164.000000 0.000000 NaN 17.000000 0.000000 0.000000 0.000000 0.000000 0.000000
75% NaN NaN 2060.000000 NaN NaN NaN NaN NaN 239.000000 230.000000 12.000000 NaN 21.000000 0.000000 0.000000 0.000000 0.000000 0.000000
max NaN NaN 3010.000000 NaN NaN NaN NaN NaN 685.000000 711.000000 1403.000000 NaN 184.000000 1215.000000 876.000000 999.000000 70.000000 1280.000000

Methods for summarizing categorical variables¶

In [10]:
united_2022['Flight Number'].mode() # Mode, most frequent value
Out[10]:
0    1561
Name: Flight Number, dtype: int64
In [11]:
united_2022['Origin Airport'].value_counts() # Counts of unique values
Out[11]:
DEN    32245
EWR    25761
ORD    24767
IAH    22561
SFO    16511
IAD    11286
LAX     9986
MCO     5680
DFW     2658
ATL     1297
PHL     1221
JFK      602
Name: Origin Airport, dtype: int64
In [12]:
united_2022['Origin Airport'].value_counts(normalize = True) # We can further normalize them by the number of rows
Out[12]:
DEN    0.208604
EWR    0.166657
ORD    0.160226
IAH    0.145955
SFO    0.106815
IAD    0.073013
LAX    0.064603
MCO    0.036746
DFW    0.017196
ATL    0.008391
PHL    0.007899
JFK    0.003895
Name: Origin Airport, dtype: float64

Summary of descriptive statistics methods¶

Method Numeric Categorical Description
count yes yes Number of non-NA observations
value_counts yes yes Number of unique observations by value
describe yes yes Set of summary statistics for Series/DataFrame
min, max yes yes (caution) Minimum and maximum values
quantile yes no Sample quantile ranging from 0 to 1
sum yes yes (caution) Sum of values
prod yes no Product of values
mean yes no Mean
median yes no Median (50% quantile)
var yes no Sample variance
std yes no Sample standard deviation
skew yes no Sample skewness (third moment)
kurt yes no Sample kurtosis (fourth moment)

Crosstabulation¶

  • When working with survey data it is often useful to perform simple crosstabulations
  • Crosstabulation (or crosstab for short) is a computation of group frequencies
  • It is usually used for working with categorical variables that have a limited number of categories
  • In pandas pd.crosstab() method is a special case of pd.pivot_table()

Crosstabulation in pandas¶

In [13]:
# Calculate crosstabulation between origin and destination airports
pd.crosstab(united_2022['Origin Airport'], united_2022['Destination Airport'])
Out[13]:
Destination Airport ABQ ALB ANC ATL AUS BDL BIL BNA BOI BOS ... SLC SMF SNA SRQ STL STT SYR TPA TUL TUS
Origin Airport
ATL 0 0 0 0 0 0 0 0 0 0 ... 0 0 0 0 0 0 0 0 0 0
DEN 267 0 154 313 585 151 300 151 300 670 ... 422 633 617 151 257 0 110 565 108 232
DFW 0 0 0 0 0 0 0 0 0 0 ... 0 0 0 0 0 0 0 0 0 0
EWR 0 0 0 586 553 0 0 346 0 1280 ... 221 150 375 290 0 153 1 1157 0 0
IAD 0 0 0 147 270 340 0 162 0 480 ... 0 88 0 2 0 151 123 400 0 0
IAH 68 0 0 23 897 86 0 112 0 464 ... 139 175 537 0 96 130 0 637 118 36
JFK 0 0 0 0 0 0 0 0 0 0 ... 0 0 0 0 0 0 0 0 0 0
LAX 0 0 0 0 111 0 0 0 0 250 ... 0 0 0 0 0 0 0 142 0 0
MCO 0 0 0 0 0 0 0 0 0 0 ... 0 0 0 0 0 0 0 0 0 0
ORD 29 199 66 205 518 314 0 141 0 643 ... 105 280 497 224 0 40 108 576 0 72
PHL 0 0 0 0 0 0 0 0 0 0 ... 0 0 0 0 0 0 0 0 0 0
SFO 81 0 0 26 421 0 0 150 0 425 ... 63 0 488 0 0 0 0 248 0 0

12 rows × 110 columns

Margins in crosstab¶

In [14]:
# It is often useful to see the proportions/percentages rather than raw counts
pd.crosstab(united_2022['Origin Airport'], united_2022['Destination Airport'], normalize = 'columns')
Out[14]:
Destination Airport ABQ ALB ANC ATL AUS BDL BIL BNA BOI BOS ... SLC SMF SNA SRQ STL STT SYR TPA TUL TUS
Origin Airport
ATL 0.000000 0.0 0.0 0.000000 0.000000 0.000000 0.0 0.000000 0.0 0.000000 ... 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000
DEN 0.600000 0.0 0.7 0.240769 0.174367 0.169473 1.0 0.142185 1.0 0.159069 ... 0.444211 0.477376 0.245426 0.226387 0.728045 0.000000 0.321637 0.151678 0.477876 0.682353
DFW 0.000000 0.0 0.0 0.000000 0.000000 0.000000 0.0 0.000000 0.0 0.000000 ... 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000
EWR 0.000000 0.0 0.0 0.450769 0.164829 0.000000 0.0 0.325800 0.0 0.303894 ... 0.232632 0.113122 0.149165 0.434783 0.000000 0.322785 0.002924 0.310604 0.000000 0.000000
IAD 0.000000 0.0 0.0 0.113077 0.080477 0.381594 0.0 0.152542 0.0 0.113960 ... 0.000000 0.066365 0.000000 0.002999 0.000000 0.318565 0.359649 0.107383 0.000000 0.000000
IAH 0.152809 0.0 0.0 0.017692 0.267362 0.096521 0.0 0.105461 0.0 0.110161 ... 0.146316 0.131976 0.213604 0.000000 0.271955 0.274262 0.000000 0.171007 0.522124 0.105882
JFK 0.000000 0.0 0.0 0.000000 0.000000 0.000000 0.0 0.000000 0.0 0.000000 ... 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000
LAX 0.000000 0.0 0.0 0.000000 0.033085 0.000000 0.0 0.000000 0.0 0.059354 ... 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000 0.038121 0.000000 0.000000
MCO 0.000000 0.0 0.0 0.000000 0.000000 0.000000 0.0 0.000000 0.0 0.000000 ... 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000
ORD 0.065169 1.0 0.3 0.157692 0.154396 0.352413 0.0 0.132768 0.0 0.152659 ... 0.110526 0.211161 0.197693 0.335832 0.000000 0.084388 0.315789 0.154631 0.000000 0.211765
PHL 0.000000 0.0 0.0 0.000000 0.000000 0.000000 0.0 0.000000 0.0 0.000000 ... 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000
SFO 0.182022 0.0 0.0 0.020000 0.125484 0.000000 0.0 0.141243 0.0 0.100902 ... 0.066316 0.000000 0.194113 0.000000 0.000000 0.000000 0.000000 0.066577 0.000000 0.000000

12 rows × 110 columns

Data visualization¶

Source: Tufte (2001), based on Marey (1885)

Data visualization in Python¶

  • As with dealing with data, Python has no in-built, 'base' plotting functionality
  • matplotlib has become the one of standard solutions
  • It is often used in combination with pandas
  • Other popular alternative include seaborn and plotnine
  • Also pandas itself has some limited plotting facilities

plotnine - ggplot for Python¶

  • plotnine implements Grammar of Graphics data visualisation scheme (Wilkinson, 2005)
  • It mimics the syntax of a well-known R library ggplot2syntax (Wickham, 2010)
  • In doing so, it makes the code (almost) seamlessly portable between the two languages

Grammar of graphics¶

  • Grammar of Graphics is a powerful conceptualization of plotting
  • Graphs are broken into multiple layers
  • Layers can be recycled across multiple plots

Structure of ggplot calls in plotnine¶

  • Creation of ggplot objects in plotline has the following structure:
ggplot(data = <DATA>) +\
    <GEOM_FUNCTION>(mapping = aes(<MAPPINGS>))
  • If the mappings are re-used across geometric objects (e.g. scatterplot and line):
ggplot(data = <DATA>, mapping = aes(<MAPPINGS>)) +\
    <GEOM_FUNCTION>() +\
    <GEOM_FUNCTION>()

Creating a ggplot in plotnine¶

In [15]:
from plotnine import *
In [16]:
origin_plot = ggplot(data = united_2022) + geom_bar(aes(x = 'Origin Airport'))
origin_plot
Out[16]:
<ggplot: (8785261968216)>

Compare to base pandas¶

In [17]:
# First we need to group dataset by 'Origin Airport' and summarize it with `size()`
united_2022_origin_grouped = united_2022.groupby(['Origin Airport']).size()
In [18]:
united_2022_origin_grouped.plot(kind = 'bar')
Out[18]:
<AxesSubplot:xlabel='Origin Airport'>

Compare to matplotlib¶

In [19]:
import matplotlib.pyplot as plt
In [20]:
# `matplotlib` is more low-level library
# plots would need more work to be 'prettified'
plt.bar(x = united_2022_origin_grouped.index, height = united_2022_origin_grouped.values)
Out[20]:
<BarContainer object of 12 artists>

Prettifying ggplot in plotnine¶

In [21]:
# Here we change default axes' labels and then apply B&W theme
origin_plot_pretty = origin_plot +\
    labs(y = 'flights') +\
    theme_bw()
origin_plot_pretty
Out[21]:
<ggplot: (8785259675203)>

Other geometric objects (geom_)¶

Method Description
geom_bar(), geom_col() Bar charts
geom_boxplot() Box and whisker plot
geom_histogram() Histogram
geom_point() Scatterplot
geom_line(), geom_path() Lines
geom_map() Geographic areas
geom_smooth() Smoothed conditional means
geom_violin() Violin plots

Writing plots out in plotnine¶

  • Output format is automatically determined from write-out file extension
  • Commonly used formats are PDF, PNG and EPS
In [22]:
origin_plot_pretty.save('../temp/q1_plot_pretty.pdf')

/home/tpaskhalis/Decrypted/Git/DS3/venv/lib/python3.8/site-packages/plotnine/ggplot.py:718: PlotnineWarning: Saving 6.4 x 4.8 in image.
/home/tpaskhalis/Decrypted/Git/DS3/venv/lib/python3.8/site-packages/plotnine/ggplot.py:719: PlotnineWarning: Filename: ../temp/q1_plot_pretty.pdf

Additional visualization materials¶

Books:

  • Hieran, Kiely. 2019. Data Visualization: A Practical Introduction. Princeton, NJ: Princeton University Press
  • Tufte, Edward. 2001. The Visual Display of Quantitative Information. 2nd ed. Cheshire, CT: Graphics Press

Online:

  • Plotnine: Grammar of Graphics for Python
  • Plotnine Documentation
  • ggplot2 Documentation

The end¶