Introduction to Data Analysis Using Pandas – Stefanie Molin

Introduction to Data Analysis Using Pandas

Stefanie Molin

Bio

• 👩‍💻 Software engineer at Bloomberg in NYC
• 🚀 Core developer of numpydoc
• ✍️ Author of Hands-On Data Analysis with Pandas (currently in its second edition; translated into Korean and Simplified Chinese)
• 🎓 BS in operations research from Columbia University
• 🎓 MS in computer science (ML specialization) from Georgia Tech

Prerequisites

• To follow along with this workshop, you will need to either configure a local environment or use a cloud solution like GitHub Codespaces. Consult the README for step-by-step setup instructions.
• In addition, you should have basic knowledge of Python and be comfortable working in Jupyter Notebooks; if not, check out the resources here to get up to speed.

Session Outline

1. Getting Started With Pandas
2. Data Wrangling
3. Data Visualization
4. Hands-On Data Analysis Lab

Let's get started

Be sure to follow along in the notebooks.

Section 1: Getting Started With Pandas

We will begin by introducing the Series, DataFrame, and Index classes, which are the basic building blocks of the pandas library, and showing how to work with them. By the end of this section, you will be able to create DataFrames and perform operations on them to inspect and filter the data.

Source

Learning Path

1. Anatomy of a DataFrame
2. Creating DataFrames
3. Inspecting the data
4. Extracting subsets
5. Calculating summary statistics

Anatomy of a DataFrame

A DataFrame is composed of one or more Series. The names of the Series form the column names, and the row labels form the Index.

import pandas as pd

meteorites = pd.read_csv('../data/Meteorite_Landings.csv', nrows=5)
meteorites

	name	id	nametype	recclass	mass (g)	fall	year	reclat	reclong	GeoLocation
0	Aachen	1	Valid	L5	21	Fell	01/01/1880 12:00:00 AM	50.77500	6.08333	(50.775, 6.08333)
1	Aarhus	2	Valid	H6	720	Fell	01/01/1951 12:00:00 AM	56.18333	10.23333	(56.18333, 10.23333)
2	Abee	6	Valid	EH4	107000	Fell	01/01/1952 12:00:00 AM	54.21667	-113.00000	(54.21667, -113.0)
3	Acapulco	10	Valid	Acapulcoite	1914	Fell	01/01/1976 12:00:00 AM	16.88333	-99.90000	(16.88333, -99.9)
4	Achiras	370	Valid	L6	780	Fell	01/01/1902 12:00:00 AM	-33.16667	-64.95000	(-33.16667, -64.95)

Source: NASA's Open Data Portal

Series:

meteorites.name

0      Aachen
1      Aarhus
2        Abee
3    Acapulco
4     Achiras
Name: name, dtype: object

Columns:

meteorites.columns

Index(['name', 'id', 'nametype', 'recclass', 'mass (g)', 'fall', 'year',
       'reclat', 'reclong', 'GeoLocation'],
      dtype='object')

Index:

meteorites.index

RangeIndex(start=0, stop=5, step=1)

Learning Path

1. Anatomy of a DataFrame
2. Creating DataFrames
3. Inspecting the data
4. Extracting subsets
5. Calculating summary statistics

Creating DataFrames

We can create DataFrames from a variety of sources such as other Python objects, flat files, webscraping, and API requests. Here, we will see just a couple of examples, but be sure to check out this page in the documentation for a complete list.

Using a flat file

import pandas as pd

meteorites = pd.read_csv('../data/Meteorite_Landings.csv')

Using data from an API

Collect the data from NASA's Open Data Portal using the Socrata Open Data API (SODA) with the requests library:

import requests

response = requests.get(
    'https://data.nasa.gov/resource/gh4g-9sfh.json',
    params={'$limit': 50_000}
)

if response.ok:
    payload = response.json()
else:
    print(f'Request was not successful and returned code: {response.status_code}.')
    payload = None

Create the DataFrame with the resulting payload:

import pandas as pd

df = pd.DataFrame(payload)
df.head(3)

	name	id	nametype	recclass	mass	fall	year	reclat	reclong	geolocation	:@computed_region_cbhk_fwbd	:@computed_region_nnqa_25f4
0	Aachen	1	Valid	L5	21	Fell	1880-01-01T00:00:00.000	50.775000	6.083330	{'latitude': '50.775', 'longitude': '6.08333'}	NaN	NaN
1	Aarhus	2	Valid	H6	720	Fell	1951-01-01T00:00:00.000	56.183330	10.233330	{'latitude': '56.18333', 'longitude': '10.23333'}	NaN	NaN
2	Abee	6	Valid	EH4	107000	Fell	1952-01-01T00:00:00.000	54.216670	-113.000000	{'latitude': '54.21667', 'longitude': '-113.0'}	NaN	NaN

Tip: df.to_csv('data.csv') writes this data to a new file called data.csv.

Learning Path

1. Anatomy of a DataFrame
2. Creating DataFrames
3. Inspecting the data
4. Extracting subsets
5. Calculating summary statistics

Inspecting the data

Now that we have some data, we need to perform an initial inspection of it. This gives us information on what the data looks like, how many rows/columns there are, and how much data we have.

Let's inspect the meteorites data.

How many rows and columns are there?

meteorites.shape

(45716, 10)

What are the column names?

meteorites.columns

Index(['name', 'id', 'nametype', 'recclass', 'mass (g)', 'fall', 'year',
       'reclat', 'reclong', 'GeoLocation'],
      dtype='object')

What type of data does each column currently hold?

meteorites.dtypes

name            object
id               int64
nametype        object
recclass        object
mass (g)       float64
fall            object
year            object
reclat         float64
reclong        float64
GeoLocation     object
dtype: object

What does the data look like?

meteorites.head()

	name	id	nametype	recclass	mass (g)	fall	year	reclat	reclong	GeoLocation
0	Aachen	1	Valid	L5	21.0	Fell	01/01/1880 12:00:00 AM	50.77500	6.08333	(50.775, 6.08333)
1	Aarhus	2	Valid	H6	720.0	Fell	01/01/1951 12:00:00 AM	56.18333	10.23333	(56.18333, 10.23333)
2	Abee	6	Valid	EH4	107000.0	Fell	01/01/1952 12:00:00 AM	54.21667	-113.00000	(54.21667, -113.0)
3	Acapulco	10	Valid	Acapulcoite	1914.0	Fell	01/01/1976 12:00:00 AM	16.88333	-99.90000	(16.88333, -99.9)
4	Achiras	370	Valid	L6	780.0	Fell	01/01/1902 12:00:00 AM	-33.16667	-64.95000	(-33.16667, -64.95)

Sometimes there may be extraneous data at the end of the file, so checking the bottom few rows is also important:

meteorites.tail()

	name	id	nametype	recclass	mass (g)	fall	year	reclat	reclong	GeoLocation
45711	Zillah 002	31356	Valid	Eucrite	172.0	Found	01/01/1990 12:00:00 AM	29.03700	17.01850	(29.037, 17.0185)
45712	Zinder	30409	Valid	Pallasite, ungrouped	46.0	Found	01/01/1999 12:00:00 AM	13.78333	8.96667	(13.78333, 8.96667)
45713	Zlin	30410	Valid	H4	3.3	Found	01/01/1939 12:00:00 AM	49.25000	17.66667	(49.25, 17.66667)
45714	Zubkovsky	31357	Valid	L6	2167.0	Found	01/01/2003 12:00:00 AM	49.78917	41.50460	(49.78917, 41.5046)
45715	Zulu Queen	30414	Valid	L3.7	200.0	Found	01/01/1976 12:00:00 AM	33.98333	-115.68333	(33.98333, -115.68333)

Get some information about the DataFrame

meteorites.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 45716 entries, 0 to 45715
Data columns (total 10 columns):
 #   Column       Non-Null Count  Dtype  
---  ------       --------------  -----  
 0   name         45716 non-null  object 
 1   id           45716 non-null  int64  
 2   nametype     45716 non-null  object 
 3   recclass     45716 non-null  object 
 4   mass (g)     45585 non-null  float64
 5   fall         45716 non-null  object 
 6   year         45425 non-null  object 
 7   reclat       38401 non-null  float64
 8   reclong      38401 non-null  float64
 9   GeoLocation  38401 non-null  object 
dtypes: float64(3), int64(1), object(6)
memory usage: 3.5+ MB

Exercise 1.1

Create a DataFrame by reading in the 2019_Yellow_Taxi_Trip_Data.csv file. Examine the first 5 rows.

Exercise 1.2

Find the dimensions (number of rows and number of columns) in the data.

Solutions

Exercise 1.1 – Create a DataFrame by reading in the 2019_Yellow_Taxi_Trip_Data.csv file. Examine the first 5 rows:

import pandas as pd

taxis = pd.read_csv('../data/2019_Yellow_Taxi_Trip_Data.csv')
taxis.head()

	vendorid	tpep_pickup_datetime	tpep_dropoff_datetime	passenger_count	trip_distance	ratecodeid	store_and_fwd_flag	pulocationid	dolocationid	payment_type	fare_amount	extra	mta_tax	tip_amount	tolls_amount	improvement_surcharge	total_amount	congestion_surcharge
0	2	2019-10-23T16:39:42.000	2019-10-23T17:14:10.000	1	7.93	1	N	138	170	1	29.5	1.0	0.5	7.98	6.12	0.3	47.90	2.5
1	1	2019-10-23T16:32:08.000	2019-10-23T16:45:26.000	1	2.00	1	N	11	26	1	10.5	1.0	0.5	0.00	0.00	0.3	12.30	0.0
2	2	2019-10-23T16:08:44.000	2019-10-23T16:21:11.000	1	1.36	1	N	163	162	1	9.5	1.0	0.5	2.00	0.00	0.3	15.80	2.5
3	2	2019-10-23T16:22:44.000	2019-10-23T16:43:26.000	1	1.00	1	N	170	163	1	13.0	1.0	0.5	4.32	0.00	0.3	21.62	2.5
4	2	2019-10-23T16:45:11.000	2019-10-23T16:58:49.000	1	1.96	1	N	163	236	1	10.5	1.0	0.5	0.50	0.00	0.3	15.30	2.5

Source: NYC Open Data collected via SODA.

Exercise 1.2 – Find the dimensions (number of rows and number of columns) in the data:

taxis.shape

(10000, 18)

Learning Path

1. Anatomy of a DataFrame
2. Creating DataFrames
3. Inspecting the data
4. Extracting subsets
5. Calculating summary statistics

Extracting subsets

A crucial part of working with DataFrames is extracting subsets of the data: finding rows that meet a certain set of criteria, isolating columns/rows of interest, etc. After narrowing down our data, we are closer to discovering insights. This section will be the backbone of many analysis tasks.

Selecting columns

We can select columns as attributes if their names would be valid Python variables:

meteorites.name

0            Aachen
1            Aarhus
2              Abee
3          Acapulco
4           Achiras
            ...    
45711    Zillah 002
45712        Zinder
45713          Zlin
45714     Zubkovsky
45715    Zulu Queen
Name: name, Length: 45716, dtype: object

If they aren't, we have to select them as keys. However, we can select multiple columns at once this way:

meteorites[['name', 'mass (g)']]

	name	mass (g)
0	Aachen	21.0
1	Aarhus	720.0
2	Abee	107000.0
3	Acapulco	1914.0
4	Achiras	780.0
...	...	...
45711	Zillah 002	172.0
45712	Zinder	46.0
45713	Zlin	3.3
45714	Zubkovsky	2167.0
45715	Zulu Queen	200.0

45716 rows × 2 columns

Selecting rows

meteorites[100:104]

	name	id	nametype	recclass	mass (g)	fall	year	reclat	reclong	GeoLocation
100	Benton	5026	Valid	LL6	2840.0	Fell	01/01/1949 12:00:00 AM	45.95000	-67.55000	(45.95, -67.55)
101	Berduc	48975	Valid	L6	270.0	Fell	01/01/2008 12:00:00 AM	-31.91000	-58.32833	(-31.91, -58.32833)
102	Béréba	5028	Valid	Eucrite-mmict	18000.0	Fell	01/01/1924 12:00:00 AM	11.65000	-3.65000	(11.65, -3.65)
103	Berlanguillas	5029	Valid	L6	1440.0	Fell	01/01/1811 12:00:00 AM	41.68333	-3.80000	(41.68333, -3.8)

Indexing

We use iloc[] to select rows and columns by their position:

meteorites.iloc[100:104, [0, 3, 4, 6]]

	name	recclass	mass (g)	year
100	Benton	LL6	2840.0	01/01/1949 12:00:00 AM
101	Berduc	L6	270.0	01/01/2008 12:00:00 AM
102	Béréba	Eucrite-mmict	18000.0	01/01/1924 12:00:00 AM
103	Berlanguillas	L6	1440.0	01/01/1811 12:00:00 AM

We use loc[] to select by name:

meteorites.loc[100:104, 'mass (g)':'year']

	mass (g)	fall	year
100	2840.0	Fell	01/01/1949 12:00:00 AM
101	270.0	Fell	01/01/2008 12:00:00 AM
102	18000.0	Fell	01/01/1924 12:00:00 AM
103	1440.0	Fell	01/01/1811 12:00:00 AM
104	960.0	Fell	01/01/2004 12:00:00 AM

Filtering with Boolean masks

A Boolean mask is a array-like structure of Boolean values – it's a way to specify which rows/columns we want to select (True) and which we don't (False).

Here's an example of a Boolean mask for meteorites weighing more than 50 grams that were found on Earth (i.e., they were not observed falling):

(meteorites['mass (g)'] > 50) & (meteorites.fall == 'Found')

0        False
1        False
2        False
3        False
4        False
         ...  
45711     True
45712    False
45713    False
45714     True
45715     True
Length: 45716, dtype: bool

Important: Take note of the syntax here. We surround each condition with parentheses, and we use bitwise operators (&, |, ~) instead of logical operators (and, or, not).

We can use a Boolean mask to select the subset of meteorites weighing more than 1 million grams (1,000 kilograms or roughly 2,205 pounds) that were observed falling:

meteorites[(meteorites['mass (g)'] > 1e6) & (meteorites.fall == 'Fell')]

	name	id	nametype	recclass	mass (g)	fall	year	reclat	reclong	GeoLocation
29	Allende	2278	Valid	CV3	2000000.0	Fell	01/01/1969 12:00:00 AM	26.96667	-105.31667	(26.96667, -105.31667)
419	Jilin	12171	Valid	H5	4000000.0	Fell	01/01/1976 12:00:00 AM	44.05000	126.16667	(44.05, 126.16667)
506	Kunya-Urgench	12379	Valid	H5	1100000.0	Fell	01/01/1998 12:00:00 AM	42.25000	59.20000	(42.25, 59.2)
707	Norton County	17922	Valid	Aubrite	1100000.0	Fell	01/01/1948 12:00:00 AM	39.68333	-99.86667	(39.68333, -99.86667)
920	Sikhote-Alin	23593	Valid	Iron, IIAB	23000000.0	Fell	01/01/1947 12:00:00 AM	46.16000	134.65333	(46.16, 134.65333)

Tip: Boolean masks can be used with loc[] and iloc[].

An alternative to this is the query() method:

meteorites.query("`mass (g)` > 1e6 and fall == 'Fell'")

	name	id	nametype	recclass	mass (g)	fall	year	reclat	reclong	GeoLocation
29	Allende	2278	Valid	CV3	2000000.0	Fell	01/01/1969 12:00:00 AM	26.96667	-105.31667	(26.96667, -105.31667)
419	Jilin	12171	Valid	H5	4000000.0	Fell	01/01/1976 12:00:00 AM	44.05000	126.16667	(44.05, 126.16667)
506	Kunya-Urgench	12379	Valid	H5	1100000.0	Fell	01/01/1998 12:00:00 AM	42.25000	59.20000	(42.25, 59.2)
707	Norton County	17922	Valid	Aubrite	1100000.0	Fell	01/01/1948 12:00:00 AM	39.68333	-99.86667	(39.68333, -99.86667)
920	Sikhote-Alin	23593	Valid	Iron, IIAB	23000000.0	Fell	01/01/1947 12:00:00 AM	46.16000	134.65333	(46.16, 134.65333)

Tip: Here, we can use both logical operators and bitwise operators.

Learning Path

1. Anatomy of a DataFrame
2. Creating DataFrames
3. Inspecting the data
4. Extracting subsets
5. Calculating summary statistics

Calculating summary statistics

In the next section of this workshop, we will discuss data cleaning for a more meaningful analysis of our datasets; however, we can already extract some interesting insights from the meteorites data by calculating summary statistics.

How many of the meteorites were found versus observed falling?

meteorites.fall.value_counts()

fall
Found    44609
Fell      1107
Name: count, dtype: int64

The Meteoritical Society states that "relict meteorites are composed mostly of terrestrial minerals, but are thought to have once been meteorites." What proportion of the data are relict meteorites? Let's verify these were all found versus observed falling:

df.value_counts(subset=['nametype', 'fall'], normalize=True)

nametype  fall 
Valid     Found    0.974145
          Fell     0.024215
Relict    Found    0.001641
Name: proportion, dtype: float64

What was the mass of the average meterorite?

meteorites['mass (g)'].mean()

np.float64(13278.078548601512)

Important: The mean isn't always the best measure of central tendency. If there are outliers in the distribution, the mean will be skewed. Here, the mean is being pulled higher by some very heavy meteorites – the distribution is right-skewed.

Taking a look at some quantiles at the extremes of the distribution shows that the mean is between the 95th and 99th percentile of the distribution, so it isn't a good measure of central tendency here:

meteorites['mass (g)'].quantile([0.01, 0.05, 0.5, 0.95, 0.99])

0.01        0.44
0.05        1.10
0.50       32.60
0.95     4000.00
0.99    50600.00
Name: mass (g), dtype: float64

A better measure in this case is the median (50th percentile), since it is robust to outliers:

meteorites['mass (g)'].median()

np.float64(32.6)

What was the mass of the heaviest meteorite?

meteorites['mass (g)'].max()

np.float64(60000000.0)

Let's extract the information on this meteorite:

meteorites.loc[meteorites['mass (g)'].idxmax()]

name                             Hoba
id                              11890
nametype                        Valid
recclass                    Iron, IVB
mass (g)                   60000000.0
fall                            Found
year           01/01/1920 12:00:00 AM
reclat                      -19.58333
reclong                      17.91667
GeoLocation     (-19.58333, 17.91667)
Name: 16392, dtype: object

Fun fact: This meteorite landed in Namibia and is a tourist attraction.

Source: Wikipedia

How many different types of meteorite classes are represented in this dataset?

meteorites.recclass.nunique()

466

Some examples:

meteorites.recclass.unique()[:14]

array(['L5', 'H6', 'EH4', 'Acapulcoite', 'L6', 'LL3-6', 'H5', 'L',
       'Diogenite-pm', 'Unknown', 'H4', 'H', 'Iron, IVA', 'CR2-an'],
      dtype=object)

Note: All fields preceded with "rec" are the values recommended by The Meteoritical Society. Check out this Wikipedia article for some information on meteorite classes.

Get some summary statistics on the data itself

We can get common summary statistics for all columns at once. By default, this will only be numeric columns, but here, we will summarize everything together:

meteorites.describe(include='all')

	name	id	nametype	recclass	mass (g)	fall	year	reclat	reclong	GeoLocation
count	45716	45716.000000	45716	45716	4.558500e+04	45716	45425	38401.000000	38401.000000	38401
unique	45716	NaN	2	466	NaN	2	266	NaN	NaN	17100
top	Aachen	NaN	Valid	L6	NaN	Found	01/01/2003 12:00:00 AM	NaN	NaN	(0.0, 0.0)
freq	1	NaN	45641	8285	NaN	44609	3323	NaN	NaN	6214
mean	NaN	26889.735104	NaN	NaN	1.327808e+04	NaN	NaN	-39.122580	61.074319	NaN
std	NaN	16860.683030	NaN	NaN	5.749889e+05	NaN	NaN	46.378511	80.647298	NaN
min	NaN	1.000000	NaN	NaN	0.000000e+00	NaN	NaN	-87.366670	-165.433330	NaN
25%	NaN	12688.750000	NaN	NaN	7.200000e+00	NaN	NaN	-76.714240	0.000000	NaN
50%	NaN	24261.500000	NaN	NaN	3.260000e+01	NaN	NaN	-71.500000	35.666670	NaN
75%	NaN	40656.750000	NaN	NaN	2.026000e+02	NaN	NaN	0.000000	157.166670	NaN
max	NaN	57458.000000	NaN	NaN	6.000000e+07	NaN	NaN	81.166670	354.473330	NaN

Important: NaN values signify missing data. For instance, the fall column contains strings, so there is no value for mean; likewise, mass (g) is numeric, so we don't have entries for the categorical summary statistics (unique, top, freq).

Check out the documentation for more descriptive statistics:

• Series
• DataFrame

Exercise 1.3

Using the data in the 2019_Yellow_Taxi_Trip_Data.csv file, calculate summary statistics for the fare_amount, tip_amount, tolls_amount, and total_amount columns.

Exercise 1.4

Isolate the fare_amount, tip_amount, tolls_amount, and total_amount for the longest trip by distance (trip_distance).

Solutions

Exercise 1.3 – Using the data in the 2019_Yellow_Taxi_Trip_Data.csv file, calculate summary statistics for the fare_amount, tip_amount, tolls_amount, and total_amount columns:

import pandas as pd

taxis = pd.read_csv('../data/2019_Yellow_Taxi_Trip_Data.csv')
taxis[['fare_amount', 'tip_amount', 'tolls_amount', 'total_amount']].describe()

	fare_amount	tip_amount	tolls_amount	total_amount
count	10000.000000	10000.000000	10000.000000	10000.000000
mean	15.106313	2.634494	0.623447	22.564659
std	13.954762	3.409800	6.437507	19.209255
min	-52.000000	0.000000	-6.120000	-65.920000
25%	7.000000	0.000000	0.000000	12.375000
50%	10.000000	2.000000	0.000000	16.300000
75%	16.000000	3.250000	0.000000	22.880000
max	176.000000	43.000000	612.000000	671.800000

Exercise 1.4 – Isolate the fare_amount, tip_amount, tolls_amount, and total_amount for the longest trip by distance (trip_distance):

taxis.loc[
    taxis.trip_distance.idxmax(), 
    ['fare_amount', 'tip_amount', 'tolls_amount', 'total_amount']
]

fare_amount      176.0
tip_amount       18.29
tolls_amount      6.12
total_amount    201.21
Name: 8338, dtype: object

Section 1 Complete 🎉

Section 2: Data Wrangling

To prepare our data for analysis, we need to perform data wrangling. In this section, we will learn how to clean and reformat data (e.g., renaming columns and fixing data type mismatches), restructure/reshape it, and enrich it (e.g., discretizing columns, calculating aggregations, and combining data sources).

Source

Learning Path

1. Data cleaning
2. Working with the index
3. Reshaping data
4. Aggregations and grouping
5. Time series

Data cleaning

In this section, we will take a look at creating, renaming, and dropping columns; type conversion; and sorting – all of which make our analysis easier. We will be working with the 2019 Yellow Taxi Trip Data provided by NYC Open Data.

import pandas as pd

taxis = pd.read_csv('../data/2019_Yellow_Taxi_Trip_Data.csv')
taxis.head()

	vendorid	tpep_pickup_datetime	tpep_dropoff_datetime	passenger_count	trip_distance	ratecodeid	store_and_fwd_flag	pulocationid	dolocationid	payment_type	fare_amount	extra	mta_tax	tip_amount	tolls_amount	improvement_surcharge	total_amount	congestion_surcharge
0	2	2019-10-23T16:39:42.000	2019-10-23T17:14:10.000	1	7.93	1	N	138	170	1	29.5	1.0	0.5	7.98	6.12	0.3	47.90	2.5
1	1	2019-10-23T16:32:08.000	2019-10-23T16:45:26.000	1	2.00	1	N	11	26	1	10.5	1.0	0.5	0.00	0.00	0.3	12.30	0.0
2	2	2019-10-23T16:08:44.000	2019-10-23T16:21:11.000	1	1.36	1	N	163	162	1	9.5	1.0	0.5	2.00	0.00	0.3	15.80	2.5
3	2	2019-10-23T16:22:44.000	2019-10-23T16:43:26.000	1	1.00	1	N	170	163	1	13.0	1.0	0.5	4.32	0.00	0.3	21.62	2.5
4	2	2019-10-23T16:45:11.000	2019-10-23T16:58:49.000	1	1.96	1	N	163	236	1	10.5	1.0	0.5	0.50	0.00	0.3	15.30	2.5

Source: NYC Open Data collected via SODA.

Dropping columns

Let's start by dropping the ID columns and the store_and_fwd_flag column, which we won't be using.

mask = taxis.columns.str.contains('id$|store_and_fwd_flag', regex=True)
columns_to_drop = taxis.columns[mask]
columns_to_drop

Index(['vendorid', 'ratecodeid', 'store_and_fwd_flag', 'pulocationid',
       'dolocationid'],
      dtype='object')

taxis = taxis.drop(columns=columns_to_drop)
taxis.head()

	tpep_pickup_datetime	tpep_dropoff_datetime	passenger_count	trip_distance	payment_type	fare_amount	extra	mta_tax	tip_amount	tolls_amount	improvement_surcharge	total_amount	congestion_surcharge
0	2019-10-23T16:39:42.000	2019-10-23T17:14:10.000	1	7.93	1	29.5	1.0	0.5	7.98	6.12	0.3	47.90	2.5
1	2019-10-23T16:32:08.000	2019-10-23T16:45:26.000	1	2.00	1	10.5	1.0	0.5	0.00	0.00	0.3	12.30	0.0
2	2019-10-23T16:08:44.000	2019-10-23T16:21:11.000	1	1.36	1	9.5	1.0	0.5	2.00	0.00	0.3	15.80	2.5
3	2019-10-23T16:22:44.000	2019-10-23T16:43:26.000	1	1.00	1	13.0	1.0	0.5	4.32	0.00	0.3	21.62	2.5
4	2019-10-23T16:45:11.000	2019-10-23T16:58:49.000	1	1.96	1	10.5	1.0	0.5	0.50	0.00	0.3	15.30	2.5

Tip: Another way to do this is to select the columns we want to keep: taxis.loc[:,~mask].

Renaming columns

Next, let's rename the datetime columns:

taxis = taxis.rename(
    columns={
        'tpep_pickup_datetime': 'pickup', 
        'tpep_dropoff_datetime': 'dropoff'
    }
)
taxis.columns

Index(['pickup', 'dropoff', 'passenger_count', 'trip_distance', 'payment_type',
       'fare_amount', 'extra', 'mta_tax', 'tip_amount', 'tolls_amount',
       'improvement_surcharge', 'total_amount', 'congestion_surcharge'],
      dtype='object')

Type conversion

Notice anything off with the data types?

taxis.dtypes

pickup                    object
dropoff                   object
passenger_count            int64
trip_distance            float64
payment_type               int64
fare_amount              float64
extra                    float64
mta_tax                  float64
tip_amount               float64
tolls_amount             float64
improvement_surcharge    float64
total_amount             float64
congestion_surcharge     float64
dtype: object

Both pickup and dropoff should be stored as datetimes. Let's fix this:

taxis[['pickup', 'dropoff']] = \
    taxis[['pickup', 'dropoff']].apply(pd.to_datetime)
taxis.dtypes

pickup                   datetime64[ns]
dropoff                  datetime64[ns]
passenger_count                   int64
trip_distance                   float64
payment_type                      int64
fare_amount                     float64
extra                           float64
mta_tax                         float64
tip_amount                      float64
tolls_amount                    float64
improvement_surcharge           float64
total_amount                    float64
congestion_surcharge            float64
dtype: object

Tip: There are other ways to perform type conversion. For numeric values, we can use the pd.to_numeric() function, and we will see the astype() method, which is a more generic method, a little later.

Creating new columns

Let's calculate the following for each row:

1. elapsed time of the trip
2. the tip percentage
3. the total taxes, tolls, fees, and surcharges
4. the average speed of the taxi

taxis = taxis.assign(
    elapsed_time=lambda x: x.dropoff - x.pickup, # 1
    cost_before_tip=lambda x: x.total_amount - x.tip_amount,
    tip_pct=lambda x: x.tip_amount / x.cost_before_tip, # 2
    fees=lambda x: x.cost_before_tip - x.fare_amount, # 3
    avg_speed=lambda x: x.trip_distance.div(
        x.elapsed_time.dt.total_seconds() / 60 / 60
    ) # 4
)

Tip: New to lambda functions? These small, anonymous functions can receive multiple arguments, but can only contain one expression (the return value). You will see these a lot in pandas code. Read more about them here.

Our new columns get added to the right:

taxis.head(2)

	pickup	dropoff	passenger_count	trip_distance	payment_type	fare_amount	extra	mta_tax	tip_amount	tolls_amount	improvement_surcharge	total_amount	congestion_surcharge	elapsed_time	cost_before_tip	tip_pct	fees	avg_speed
0	2019-10-23 16:39:42	2019-10-23 17:14:10	1	7.93	1	29.5	1.0	0.5	7.98	6.12	0.3	47.9	2.5	0 days 00:34:28	39.92	0.1999	10.42	13.804642
1	2019-10-23 16:32:08	2019-10-23 16:45:26	1	2.00	1	10.5	1.0	0.5	0.00	0.00	0.3	12.3	0.0	0 days 00:13:18	12.30	0.0000	1.80	9.022556

Some things to note:

• We used lambda functions to 1) avoid typing taxis repeatedly and 2) be able to access the cost_before_tip and elapsed_time columns in the same method that we create them.
• To create a single new column, we can also use df['new_col'] = <values>.

Sorting by values

We can use the sort_values() method to sort based on any number of columns:

taxis.sort_values(['passenger_count', 'pickup'], ascending=[False, True]).head()

	pickup	dropoff	passenger_count	trip_distance	payment_type	fare_amount	extra	mta_tax	tip_amount	tolls_amount	improvement_surcharge	total_amount	congestion_surcharge	elapsed_time	cost_before_tip	tip_pct	fees	avg_speed
5997	2019-10-23 15:55:19	2019-10-23 16:08:25	6	1.58	2	10.0	1.0	0.5	0.0	0.0	0.3	14.3	2.5	0 days 00:13:06	14.3	0.000000	4.3	7.236641
443	2019-10-23 15:56:59	2019-10-23 16:04:33	6	1.46	2	7.5	1.0	0.5	0.0	0.0	0.3	11.8	2.5	0 days 00:07:34	11.8	0.000000	4.3	11.577093
8722	2019-10-23 15:57:33	2019-10-23 16:03:34	6	0.62	1	5.5	1.0	0.5	0.7	0.0	0.3	10.5	2.5	0 days 00:06:01	9.8	0.071429	4.3	6.182825
4198	2019-10-23 15:57:38	2019-10-23 16:05:07	6	1.18	1	7.0	1.0	0.5	1.0	0.0	0.3	12.3	2.5	0 days 00:07:29	11.3	0.088496	4.3	9.461024
8238	2019-10-23 15:58:31	2019-10-23 16:29:29	6	3.23	2	19.5	1.0	0.5	0.0	0.0	0.3	23.8	2.5	0 days 00:30:58	23.8	0.000000	4.3	6.258342

To pick out the largest/smallest rows, use nlargest() / nsmallest() instead. Looking at the 3 trips with the longest elapsed time, we see some possible data integrity issues:

taxis.nlargest(3, 'elapsed_time')

	pickup	dropoff	passenger_count	trip_distance	payment_type	fare_amount	extra	mta_tax	tip_amount	tolls_amount	improvement_surcharge	total_amount	congestion_surcharge	elapsed_time	cost_before_tip	tip_pct	fees	avg_speed
7576	2019-10-23 16:52:51	2019-10-24 16:51:44	1	3.75	1	17.5	1.0	0.5	0.0	0.0	0.3	21.8	2.5	0 days 23:58:53	21.8	0.0	4.3	0.156371
6902	2019-10-23 16:51:42	2019-10-24 16:50:22	1	11.19	2	39.5	1.0	0.5	0.0	0.0	0.3	41.3	0.0	0 days 23:58:40	41.3	0.0	1.8	0.466682
4975	2019-10-23 16:18:51	2019-10-24 16:17:30	1	0.70	2	7.0	1.0	0.5	0.0	0.0	0.3	11.3	2.5	0 days 23:58:39	11.3	0.0	4.3	0.029194

Exercise 2.1

Read in the meteorite data from the Meteorite_Landings.csv file, rename the mass (g) column to mass, and drop all the latitude and longitude columns. Sort the result by mass in descending order.

Solution

import pandas as pd

meteorites = pd.read_csv('../data/Meteorite_Landings.csv')
meteorites = meteorites\
    .rename(columns={'mass (g)': 'mass'})\
    .drop(columns=meteorites.columns[-3:])\
    .sort_values('mass', ascending=False)
meteorites.head()

	name	id	nametype	recclass	mass	fall	year
16392	Hoba	11890	Valid	Iron, IVB	60000000.0	Found	01/01/1920 12:00:00 AM
5373	Cape York	5262	Valid	Iron, IIIAB	58200000.0	Found	01/01/1818 12:00:00 AM
5365	Campo del Cielo	5247	Valid	Iron, IAB-MG	50000000.0	Found	12/22/1575 12:00:00 AM
5370	Canyon Diablo	5257	Valid	Iron, IAB-MG	30000000.0	Found	01/01/1891 12:00:00 AM
3455	Armanty	2335	Valid	Iron, IIIE	28000000.0	Found	01/01/1898 12:00:00 AM

Learning Path

1. Data cleaning
2. Working with the index
3. Reshaping data
4. Aggregations and grouping
5. Time series

Working with the index

So far, we haven't really worked with the index because it's just been a row number; however, we can change the values we have in the index to access additional features of the pandas library.

Setting and sorting the index

Currently, we have a RangeIndex, but we can switch to a DatetimeIndex by specifying a datetime column when calling set_index():

taxis = taxis.set_index('pickup')
taxis.head(3)

	dropoff	passenger_count	trip_distance	payment_type	fare_amount	extra	mta_tax	tip_amount	tolls_amount	improvement_surcharge	total_amount	congestion_surcharge	elapsed_time	cost_before_tip	tip_pct	fees	avg_speed
pickup
2019-10-23 16:39:42	2019-10-23 17:14:10	1	7.93	1	29.5	1.0	0.5	7.98	6.12	0.3	47.9	2.5	0 days 00:34:28	39.92	0.199900	10.42	13.804642
2019-10-23 16:32:08	2019-10-23 16:45:26	1	2.00	1	10.5	1.0	0.5	0.00	0.00	0.3	12.3	0.0	0 days 00:13:18	12.30	0.000000	1.80	9.022556
2019-10-23 16:08:44	2019-10-23 16:21:11	1	1.36	1	9.5	1.0	0.5	2.00	0.00	0.3	15.8	2.5	0 days 00:12:27	13.80	0.144928	4.30	6.554217

Since we have a sample of the full dataset, let's sort the index to order by pickup time:

taxis = taxis.sort_index()

Tip: taxis.sort_index(axis=1) will sort the columns by name. The axis parameter is present throughout the pandas library: axis=0 targets rows and axis=1 targets columns.

We can now select ranges from our data based on the datetime the same way we did with row numbers:

taxis['2019-10-23 07:45':'2019-10-23 08']

	dropoff	passenger_count	trip_distance	payment_type	fare_amount	extra	mta_tax	tip_amount	tolls_amount	improvement_surcharge	total_amount	congestion_surcharge	elapsed_time	cost_before_tip	tip_pct	fees	avg_speed
pickup
2019-10-23 07:48:58	2019-10-23 07:52:09	1	0.67	2	4.5	1.0	0.5	0.0	0.0	0.3	8.8	2.5	0 days 00:03:11	8.8	0.000000	4.3	12.628272
2019-10-23 08:02:09	2019-10-24 07:42:32	1	8.38	1	32.0	1.0	0.5	5.5	0.0	0.3	41.8	2.5	0 days 23:40:23	36.3	0.151515	4.3	0.353989
2019-10-23 08:18:47	2019-10-23 08:36:05	1	2.39	2	12.5	1.0	0.5	0.0	0.0	0.3	16.8	2.5	0 days 00:17:18	16.8	0.000000	4.3	8.289017

When not specifying a range, we use loc[]:

taxis.loc['2019-10-23 08']

	dropoff	passenger_count	trip_distance	payment_type	fare_amount	extra	mta_tax	tip_amount	tolls_amount	improvement_surcharge	total_amount	congestion_surcharge	elapsed_time	cost_before_tip	tip_pct	fees	avg_speed
pickup
2019-10-23 08:02:09	2019-10-24 07:42:32	1	8.38	1	32.0	1.0	0.5	5.5	0.0	0.3	41.8	2.5	0 days 23:40:23	36.3	0.151515	4.3	0.353989
2019-10-23 08:18:47	2019-10-23 08:36:05	1	2.39	2	12.5	1.0	0.5	0.0	0.0	0.3	16.8	2.5	0 days 00:17:18	16.8	0.000000	4.3	8.289017

Resetting the index

We will be working with time series later this section, but sometimes we want to reset our index to row numbers and restore the columns. We can make pickup a column again with the reset_index() method:

taxis = taxis.reset_index()
taxis.head()

	pickup	dropoff	passenger_count	trip_distance	payment_type	fare_amount	extra	mta_tax	tip_amount	tolls_amount	improvement_surcharge	total_amount	congestion_surcharge	elapsed_time	cost_before_tip	tip_pct	fees	avg_speed
0	2019-10-23 07:05:34	2019-10-23 08:03:16	3	14.68	1	50.0	1.0	0.5	4.0	0.0	0.3	55.8	0.0	0 days 00:57:42	51.8	0.077220	1.8	15.265165
1	2019-10-23 07:48:58	2019-10-23 07:52:09	1	0.67	2	4.5	1.0	0.5	0.0	0.0	0.3	8.8	2.5	0 days 00:03:11	8.8	0.000000	4.3	12.628272
2	2019-10-23 08:02:09	2019-10-24 07:42:32	1	8.38	1	32.0	1.0	0.5	5.5	0.0	0.3	41.8	2.5	0 days 23:40:23	36.3	0.151515	4.3	0.353989
3	2019-10-23 08:18:47	2019-10-23 08:36:05	1	2.39	2	12.5	1.0	0.5	0.0	0.0	0.3	16.8	2.5	0 days 00:17:18	16.8	0.000000	4.3	8.289017
4	2019-10-23 09:27:16	2019-10-23 09:33:13	2	1.11	2	6.0	1.0	0.5	0.0	0.0	0.3	7.8	0.0	0 days 00:05:57	7.8	0.000000	1.8	11.193277

Exercise 2.2

Using the meteorite data from the Meteorite_Landings.csv file, update the year column to only contain the year, convert it to a numeric data type, and create a new column indicating whether the meteorite was observed falling before 1970. Set the index to the id column and extract all the rows with IDs between 10,036 and 10,040 (inclusive) with loc[].

Hint 1: Use year.str.slice() to grab a substring.

Hint 2: Make sure to sort the index before using loc[] to select the range.

Bonus: There's a data entry error in the year column. Can you find it? (Don't spend too much time on this.)

Solution

import pandas as pd

meteorites = pd.read_csv('../data/Meteorite_Landings.csv').assign(
    year=lambda x: pd.to_numeric(x.year.str.slice(6, 10)),
    pre_1970=lambda x: (x.fall == 'Fell') & (x.year < 1970)
).set_index('id')
meteorites.sort_index().loc[10_036:10_040]

	name	nametype	recclass	mass (g)	fall	year	reclat	reclong	GeoLocation	pre_1970
id
10036	Enigma	Valid	H4	94.0	Found	1967.0	31.33333	-82.31667	(31.33333, -82.31667)	False
10037	Enon	Valid	Iron, ungrouped	763.0	Found	1883.0	39.86667	-83.95000	(39.86667, -83.95)	False
10038	Enshi	Valid	H5	8000.0	Fell	1974.0	30.30000	109.50000	(30.3, 109.5)	False
10039	Ensisheim	Valid	LL6	127000.0	Fell	1491.0	47.86667	7.35000	(47.86667, 7.35)	True

Note: The pd.to_datetime() function is another option here; however, it will only be able to convert dates within the supported bounds (between pd.Timestamp.min and pd.Timestamp.max), which will cause some entries that do have a year to be marked as not having one. More information can be found in the pandas documentation here. For reference, this is how the conversion can be done:

pd.to_datetime(
    meteorites.year,
    errors='coerce',  # anything that can't be converted will be NaT (null)
    format='%m/%d/%Y %I:%M:%S %p'  # the format the datetimes are currently in
)

Bonus: There's a data entry error in the year column. Can you find it?

meteorites.year.describe()

count    45425.000000
mean      1991.828817
std         25.052766
min        860.000000
25%       1987.000000
50%       1998.000000
75%       2003.000000
max       2101.000000
Name: year, dtype: float64

There's a meteorite that was reportedly found in the future:

meteorites.query(f'year > {pd.Timestamp("today").year}')

	name	nametype	recclass	mass (g)	fall	year	reclat	reclong	GeoLocation	pre_1970
id
57150	Northwest Africa 7701	Valid	CK6	55.0	Found	2101.0	0.0	0.0	(0.0, 0.0)	False

Oops! This meteorite actually was found in 2010 (more information here).

Learning Path

1. Data cleaning
2. Working with the index
3. Reshaping data
4. Aggregations and grouping
5. Time series

Reshaping data

The taxi dataset we have be working with is in a format conducive to an analysis. This isn't always the case. Let's now take a look at the TSA traveler throughput data, which compares 2021 throughput to the same day in 2020 and 2019:

tsa = pd.read_csv('../data/tsa_passenger_throughput.csv', parse_dates=['Date'])
tsa.head()

	Date	2021 Traveler Throughput	2020 Traveler Throughput	2019 Traveler Throughput
0	2021-05-14	1716561.0	250467	2664549
1	2021-05-13	1743515.0	234928	2611324
2	2021-05-12	1424664.0	176667	2343675
3	2021-05-11	1315493.0	163205	2191387
4	2021-05-10	1657722.0	215645	2512315

Source: TSA.gov

First, we will lowercase the column names and take the first word (e.g., 2021 for 2021 Traveler Throughput) to make this easier to work with:

tsa = tsa.rename(columns=lambda x: x.lower().split()[0])
tsa.head()

	date	2021	2020	2019
0	2021-05-14	1716561.0	250467	2664549
1	2021-05-13	1743515.0	234928	2611324
2	2021-05-12	1424664.0	176667	2343675
3	2021-05-11	1315493.0	163205	2191387
4	2021-05-10	1657722.0	215645	2512315

Now, we can work on reshaping it into two columns: the date and the traveler throughput from 2019 through 2021.

Starting with the long-format data below, we want to melt it into wide-format data so that we can look at the evolution of the throughput over time:

from utils import highlight_long_format

colors = {'2021': 'pink', '2020': 'skyblue', '2019': 'lightgreen'}
highlight_long_format(tsa.head(2), colors)

	date	2021	2020	2019
0	2021-05-14 00:00:00	1716561.000000	250467	2664549
1	2021-05-13 00:00:00	1743515.000000	234928	2611324

Note that the two rows above contain the same data as the six rows below:

from utils import highlight_wide_format

highlight_wide_format(tsa.head(2), colors)

	date	year	travelers
0	2021-05-14 00:00:00	2021	1716561.000000
1	2021-05-13 00:00:00	2021	1743515.000000
2	2020-05-14 00:00:00	2020	250467.000000
3	2020-05-13 00:00:00	2020	234928.000000
4	2019-05-14 00:00:00	2019	2664549.000000
5	2019-05-13 00:00:00	2019	2611324.000000

Let's work on making this transformation.

Melting

Melting helps convert our data into long format. Now, we have all the traveler throughput numbers in a single column:

tsa_melted = tsa.melt(
    id_vars='date', # column that uniquely identifies a row (can be multiple)
    var_name='year', # name for the new column created by melting
    value_name='travelers' # name for new column containing values from melted columns
)
tsa_melted.sample(5, random_state=1) # show some random entries

	date	year	travelers
974	2020-09-12	2019	1879822.0
435	2021-03-05	2020	2198517.0
1029	2020-07-19	2019	2727355.0
680	2020-07-03	2020	718988.0
867	2020-12-28	2019	2500396.0

To convert this into a time series of traveler throughput, we need to replace the year in the date column with the one in the year column. Otherwise, we are marking prior years' numbers with the wrong year.

tsa_melted = tsa_melted.assign(
    date=lambda x: pd.to_datetime(x.year + x.date.dt.strftime('-%m-%d'))
)
tsa_melted.sample(5, random_state=1)

	date	year	travelers
974	2019-09-12	2019	1879822.0
435	2020-03-05	2020	2198517.0
1029	2019-07-19	2019	2727355.0
680	2020-07-03	2020	718988.0
867	2019-12-28	2019	2500396.0

This leaves us with some null values (the dates that aren't present in the dataset):

tsa_melted.sort_values('date').tail(3)

	date	year	travelers
136	2021-12-29	2021	NaN
135	2021-12-30	2021	NaN
134	2021-12-31	2021	NaN

These can be dropped with the dropna() method:

tsa_melted = tsa_melted.dropna()
tsa_melted.sort_values('date').tail(3)

	date	year	travelers
2	2021-05-12	2021	1424664.0
1	2021-05-13	2021	1743515.0
0	2021-05-14	2021	1716561.0

Pivoting

Using the melted data, we can pivot the data to compare TSA traveler throughput on specific days across years:

tsa_pivoted = tsa_melted\
    .query('date.dt.month == 3 and date.dt.day <= 10')\
    .assign(day_in_march=lambda x: x.date.dt.day)\
    .pivot(index='year', columns='day_in_march', values='travelers')
tsa_pivoted

day_in_march	1	2	3	4	5	6	7	8	9	10
year
2019	2257920.0	1979558.0	2143619.0	2402692.0	2543689.0	2156262.0	2485430.0	2378673.0	2122898.0	2187298.0
2020	2089641.0	1736393.0	1877401.0	2130015.0	2198517.0	1844811.0	2119867.0	1909363.0	1617220.0	1702686.0
2021	1049692.0	744812.0	826924.0	1107534.0	1168734.0	992406.0	1278557.0	1119303.0	825745.0	974221.0

Important: We aren't covering the unstack() and stack() methods, which are additional ways to pivot and melt, respectively. These come in handy when we have a multi-level index (e.g., if we ran set_index() with more than one column). More information can be found here.

Transposing

The T attribute provides a quick way to flip rows and columns.

tsa_pivoted.T

year	2019	2020	2021
day_in_march
1	2257920.0	2089641.0	1049692.0
2	1979558.0	1736393.0	744812.0
3	2143619.0	1877401.0	826924.0
4	2402692.0	2130015.0	1107534.0
5	2543689.0	2198517.0	1168734.0
6	2156262.0	1844811.0	992406.0
7	2485430.0	2119867.0	1278557.0
8	2378673.0	1909363.0	1119303.0
9	2122898.0	1617220.0	825745.0
10	2187298.0	1702686.0	974221.0

Merging

We typically observe changes in air travel around the holidays, so adding information about the dates in the TSA dataset provides more context. The holidays.csv file contains a few major holidays in the United States:

holidays = pd.read_csv('../data/holidays.csv', parse_dates=True, index_col='date')
holidays.loc['2019']

	holiday
date
2019-01-01	New Year's Day
2019-05-27	Memorial Day
2019-07-04	July 4th
2019-09-02	Labor Day
2019-11-28	Thanksgiving
2019-12-24	Christmas Eve
2019-12-25	Christmas Day
2019-12-31	New Year's Eve

Merging the holidays with the TSA traveler throughput data will provide more context for our analysis:

tsa_melted_holidays = tsa_melted\
    .merge(holidays, left_on='date', right_index=True, how='left')\
    .sort_values('date')
tsa_melted_holidays.head()

	date	year	travelers	holiday
863	2019-01-01	2019	2126398.0	New Year's Day
862	2019-01-02	2019	2345103.0	NaN
861	2019-01-03	2019	2202111.0	NaN
860	2019-01-04	2019	2150571.0	NaN
859	2019-01-05	2019	1975947.0	NaN

Tip: There are many parameters for this method, so be sure to check out the documentation. To append rows, take a look at the pd.concat() function.

We can take this a step further by marking a few days before and after each holiday as part of the holiday. This would make it easier to compare holiday travel across years and look for any uptick in travel around the holidays:

tsa_melted_holiday_travel = tsa_melted_holidays.assign(
    holiday=lambda x:
        x.holiday\
            .ffill(limit=1)\
            .bfill(limit=2)
)

Tip: Check out the fillna() method documentation for additional functionality for replacing NA/NaN values.

Notice that we now have values for the day after each holiday and the two days prior. Thanksgiving in 2019 was on November 28th, so the 26th, 27th, and 29th were filled. Since we are only replacing null values, we don't override Christmas Day with the forward fill of Christmas Eve:

tsa_melted_holiday_travel.query(
    'year == "2019" and '
    '(holiday == "Thanksgiving" or holiday.str.contains("Christmas"))'
)

	date	year	travelers	holiday
899	2019-11-26	2019	1591158.0	Thanksgiving
898	2019-11-27	2019	1968137.0	Thanksgiving
897	2019-11-28	2019	2648268.0	Thanksgiving
896	2019-11-29	2019	2882915.0	Thanksgiving
873	2019-12-22	2019	1981433.0	Christmas Eve
872	2019-12-23	2019	1937235.0	Christmas Eve
871	2019-12-24	2019	2552194.0	Christmas Eve
870	2019-12-25	2019	2582580.0	Christmas Day
869	2019-12-26	2019	2470786.0	Christmas Day

Learning Path

1. Data cleaning
2. Working with the index
3. Reshaping data
4. Aggregations and grouping
5. Time series

Aggregations and grouping

After reshaping and cleaning our data, we can perform aggregations to summarize it in a variety of ways. In this section, we will explore using pivot tables, crosstabs, and group by operations to aggregate the data.

Pivot tables

We can build a pivot table to compare holiday travel across the years in our dataset:

tsa_melted_holiday_travel.pivot_table(
    index='year', columns='holiday', sort=False,
    values='travelers', aggfunc='sum'
)

holiday	New Year's Day	Memorial Day	July 4th	Labor Day	Thanksgiving	Christmas Eve	Christmas Day	New Year's Eve
year
2019	4471501.0	9720691.0	9414228.0	8314811.0	9090478.0	6470862.0	5053366.0	6535464.0
2020	4490388.0	1126253.0	2682541.0	2993653.0	3364358.0	3029810.0	1745242.0	3057449.0
2021	1998871.0	NaN	NaN	NaN	NaN	NaN	NaN	NaN

We can use the pct_change() method on this result to see which holiday travel periods saw the biggest change in travel:

tsa_melted_holiday_travel.pivot_table(
    index='year', columns='holiday', sort=False,
    values='travelers', aggfunc='sum'
).pct_change(fill_method=None)

holiday	New Year's Day	Memorial Day	July 4th	Labor Day	Thanksgiving	Christmas Eve	Christmas Day	New Year's Eve
year
2019	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN
2020	0.004224	-0.884139	-0.715055	-0.639961	-0.629903	-0.531776	-0.654638	-0.532176
2021	-0.554856	NaN	NaN	NaN	NaN	NaN	NaN	NaN

Let's make one last pivot table with column and row subtotals, along with some formatting improvements. First, we set a display option for all floats:

pd.set_option('display.float_format', '{:,.0f}'.format)

Next, we group together Christmas Eve and Christmas Day, likewise for New Year's Eve and New Year's Day (handling for the change in year), and create the pivot table:

import numpy as np

pivot_table = tsa_melted_holiday_travel.assign(
    year=lambda x: np.where(
        x.holiday == "New Year's Day", pd.to_numeric(x.year) - 1, x.year
    ).astype(str),
    holiday=lambda x: np.where(
        x.holiday.str.contains('Christmas|New Year', regex=True), 
        x.holiday.str.replace('Day|Eve', '', regex=True).str.strip(), 
        x.holiday
    )
).pivot_table(
    index='year', columns='holiday', sort=False,
    values='travelers', aggfunc='sum', 
    margins=True, margins_name='Total'
)
# reorder columns by order in the year
pivot_table.insert(5, "New Year's", pivot_table.pop("New Year's"))
pivot_table

holiday	Memorial Day	July 4th	Labor Day	Thanksgiving	Christmas	New Year's	Total
year
2018	NaN	NaN	NaN	NaN	NaN	4,471,501	4,471,501
2019	9,720,691	9,414,228	8,314,811	9,090,478	11,524,228	11,025,852	59,090,288
2020	1,126,253	2,682,541	2,993,653	3,364,358	4,775,052	5,056,320	19,998,177
Total	10,846,944	12,096,769	11,308,464	12,454,836	16,299,280	20,553,673	83,559,966

Before moving on, let's reset the display option:

pd.reset_option('display.float_format')

Tip: Read more about options in the documentation here.

Exercise 2.3

Using the meteorite data from the Meteorite_Landings.csv file, create a pivot table that shows both the number of meteorites and the 95th percentile of meteorite mass for those that were found versus observed falling per year from 2005 through 2009 (inclusive). Hint: Be sure to convert the year column to a number as we did in the previous exercise.

Solution

import pandas as pd

meteorites = pd.read_csv('../data/Meteorite_Landings.csv').assign(
    year=lambda x: pd.to_numeric(x.year.str.slice(6, 10))
)
meteorites.query('year.between(2005, 2009)').pivot_table(
    index='year', columns='fall', values='mass (g)', 
    aggfunc=['count', lambda x: x.quantile(0.95)]
).rename(columns={'<lambda>': '95th percentile'})

	count		95th percentile
fall	Fell	Found	Fell	Found
year
2005.0	NaN	874.0	NaN	4500.00
2006.0	5.0	2450.0	25008.0	1600.50
2007.0	8.0	1181.0	89675.0	1126.90
2008.0	9.0	948.0	106000.0	2274.80
2009.0	5.0	1492.0	8333.4	1397.25

Crosstabs

The pd.crosstab() function provides an easy way to create a frequency table. Here, we count the number of low-, medium-, and high-volume travel days per year, using the pd.cut() function to create three travel volume bins of equal width:

pd.crosstab(
    index=pd.cut(
        tsa_melted_holiday_travel.travelers, 
        bins=3, labels=['low', 'medium', 'high']
    ),
    columns=tsa_melted_holiday_travel.year,
    rownames=['travel_volume']
)

year	2019	2020	2021
travel_volume
low	0	277	54
medium	42	44	80
high	323	44	0

Tip: The pd.cut() function can also be used to specify custom bin ranges. For equal-sized bins based on quantiles, use the pd.qcut() function instead.

Note that the pd.crosstab() function supports other aggregations provided you pass in the data to aggregate as values and specify the aggregation with aggfunc. You can also add subtotals and normalize the data. See the documentation for more information.

Group by operations

Rather than perform aggregations, like mean() or describe(), on the full dataset at once, we can perform these calculations per group by first calling groupby():

tsa_melted_holiday_travel.groupby('year').describe(include=np.number)

	travelers
	count	mean	std	min	25%	50%	75%	max
year
2019	365.0	2.309482e+06	285061.490784	1534386.0	2091116.0	2358007.0	2538384.00	2882915.0
2020	365.0	8.818674e+05	639775.194297	87534.0	507129.0	718310.0	983745.00	2507588.0
2021	134.0	1.112632e+06	338040.673782	468933.0	807156.0	1117391.0	1409377.75	1743515.0

Groups can also be used to perform separate calculations per subset of the data. For example, we can find the highest-volume travel day per year using rank():

tsa_melted_holiday_travel.assign(
    travel_volume_rank=lambda x: x.groupby('year').travelers.rank(ascending=False)
).sort_values(['travel_volume_rank', 'year']).head(3)

	date	year	travelers	holiday	travel_volume_rank
896	2019-11-29	2019	2882915.0	Thanksgiving	1.0
456	2020-02-12	2020	2507588.0	NaN	1.0
1	2021-05-13	2021	1743515.0	NaN	1.0

The previous two examples called a single method on the grouped data, but using the agg() method we can specify any number of them:

tsa_melted_holiday_travel.assign(
    holiday_travelers=lambda x: np.where(~x.holiday.isna(), x.travelers, np.nan),
    non_holiday_travelers=lambda x: np.where(x.holiday.isna(), x.travelers, np.nan),
    year=lambda x: pd.to_numeric(x.year)
).select_dtypes(include='number').groupby('year').agg(['mean', 'std'])

	travelers		holiday_travelers		non_holiday_travelers
	mean	std	mean	std	mean	std
year
2019	2.309482e+06	285061.490784	2.271977e+06	303021.675751	2.312359e+06	283906.226598
2020	8.818674e+05	639775.194297	8.649882e+05	489938.240989	8.831619e+05	650399.772930
2021	1.112632e+06	338040.673782	9.994355e+05	273573.249680	1.114347e+06	339479.298658

Tip: The select_dtypes() method makes it possible to select columns by their data type. We can specify the data types to exclude and/or include.

In addition, we can specify which aggregations to perform on each column:

tsa_melted_holiday_travel.assign(
    holiday_travelers=lambda x: np.where(~x.holiday.isna(), x.travelers, np.nan),
    non_holiday_travelers=lambda x: np.where(x.holiday.isna(), x.travelers, np.nan)
).groupby('year').agg({
    'holiday_travelers': ['mean', 'std'], 
    'holiday': ['nunique', 'count']
})

	holiday_travelers		holiday
	mean	std	nunique	count
year
2019	2.271977e+06	303021.675751	8	26
2020	8.649882e+05	489938.240989	8	26
2021	9.994355e+05	273573.249680	1	2

We are only scratching the surface; some additional functionalities to be aware of include the following:

• We can group by multiple columns – this creates a hierarchical index.
• Groups can be excluded from calculations with the filter() method.
• We can group on content in the index using the level or name parameters e.g., groupby(level=0) or groupby(name='year').
• We can group by date ranges if we use a pd.Grouper() object.

Be sure to check out the documentation for more details.

Exercise 2.4

Using the meteorite data from the Meteorite_Landings.csv file, compare summary statistics of the mass column for the meteorites that were found versus observed falling.

Solution

import pandas as pd

meteorites = pd.read_csv('../data/Meteorite_Landings.csv')
meteorites.groupby('fall')['mass (g)'].describe()

	count	mean	std	min	25%	50%	75%	max
fall
Fell	1075.0	47070.715023	717067.125826	0.1	686.00	2800.0	10450.0	23000000.0
Found	44510.0	12461.922983	571105.752311	0.0	6.94	30.5	178.0	60000000.0

Learning Path

1. Data cleaning
2. Working with the index
3. Reshaping data
4. Aggregations and grouping
5. Time series

Time series

When working with time series data, pandas provides us with additional functionality to not just compare the observations in our dataset, but to use their relationship in time to analyze the data. In this section, we will see a few such operations for selecting date/time ranges, calculating changes over time, performing window calculations, and resampling the data to different date/time intervals.

Selecting based on date and time

Let's switch back to the taxis dataset, which has timestamps of pickups and dropoffs. First, we will set the dropoff column as the index and sort the data:

taxis = taxis.set_index('dropoff').sort_index()

We saw earlier that we can slice on the datetimes:

taxis['2019-10-24 12':'2019-10-24 13']

	pickup	passenger_count	trip_distance	payment_type	fare_amount	extra	mta_tax	tip_amount	tolls_amount	improvement_surcharge	total_amount	congestion_surcharge	elapsed_time	cost_before_tip	tip_pct	fees	avg_speed
dropoff
2019-10-24 12:30:08	2019-10-23 13:25:42	4	0.76	2	5.0	1.0	0.5	0.00	0.0	0.3	9.30	2.5	0 days 23:04:26	9.3	0.0	4.3	0.032938
2019-10-24 12:42:01	2019-10-23 13:34:03	2	1.58	1	7.5	1.0	0.5	2.36	0.0	0.3	14.16	2.5	0 days 23:07:58	11.8	0.2	4.3	0.068301

We can also represent this range with shorthand. Note that we must use loc[] here:

taxis.loc['2019-10-24 12']

	pickup	passenger_count	trip_distance	payment_type	fare_amount	extra	mta_tax	tip_amount	tolls_amount	improvement_surcharge	total_amount	congestion_surcharge	elapsed_time	cost_before_tip	tip_pct	fees	avg_speed
dropoff
2019-10-24 12:30:08	2019-10-23 13:25:42	4	0.76	2	5.0	1.0	0.5	0.00	0.0	0.3	9.30	2.5	0 days 23:04:26	9.3	0.0	4.3	0.032938
2019-10-24 12:42:01	2019-10-23 13:34:03	2	1.58	1	7.5	1.0	0.5	2.36	0.0	0.3	14.16	2.5	0 days 23:07:58	11.8	0.2	4.3	0.068301

However, if we want to look at this time range across days, we need another strategy.

We can pull out the dropoffs that happened between a certain time range on any day with the between_time() method:

taxis.between_time('12:00', '13:00')

	pickup	passenger_count	trip_distance	payment_type	fare_amount	extra	mta_tax	tip_amount	tolls_amount	improvement_surcharge	total_amount	congestion_surcharge	elapsed_time	cost_before_tip	tip_pct	fees	avg_speed
dropoff
2019-10-23 12:53:49	2019-10-23 12:35:27	5	2.49	1	13.5	1.0	0.5	2.20	0.0	0.3	20.00	2.5	0 days 00:18:22	17.8	0.123596	4.3	8.134301
2019-10-24 12:30:08	2019-10-23 13:25:42	4	0.76	2	5.0	1.0	0.5	0.00	0.0	0.3	9.30	2.5	0 days 23:04:26	9.3	0.000000	4.3	0.032938
2019-10-24 12:42:01	2019-10-23 13:34:03	2	1.58	1	7.5	1.0	0.5	2.36	0.0	0.3	14.16	2.5	0 days 23:07:58	11.8	0.200000	4.3	0.068301

Tip: The at_time() method can be used to extract all entries at a given time (e.g., 12:35:27).

For the rest of this section, we will be working with the TSA traveler throughput data. Let's start by setting the index to the date column:

tsa_melted_holiday_travel = tsa_melted_holiday_travel.set_index('date')

Calculating change over time

tsa_melted_holiday_travel.loc['2020'].assign(
    one_day_change=lambda x: x.travelers.diff(),
    seven_day_change=lambda x: x.travelers.diff(7),
).head(10)

	year	travelers	holiday	one_day_change	seven_day_change
date
2020-01-01	2020	2311732.0	New Year's Day	NaN	NaN
2020-01-02	2020	2178656.0	New Year's Day	-133076.0	NaN
2020-01-03	2020	2422272.0	NaN	243616.0	NaN
2020-01-04	2020	2210542.0	NaN	-211730.0	NaN
2020-01-05	2020	1806480.0	NaN	-404062.0	NaN
2020-01-06	2020	1815040.0	NaN	8560.0	NaN
2020-01-07	2020	2034472.0	NaN	219432.0	NaN
2020-01-08	2020	2072543.0	NaN	38071.0	-239189.0
2020-01-09	2020	1687974.0	NaN	-384569.0	-490682.0
2020-01-10	2020	2183734.0	NaN	495760.0	-238538.0

Tip: To perform operations other than subtraction, take a look at the shift() method. It also makes it possible to perform operations across columns.

Resampling

We can use resampling to aggregate time series data to a new frequency:

tsa_melted_holiday_travel['2019':'2021-Q1'].select_dtypes(include='number')\
    .resample('QE').agg(['sum', 'mean', 'std'])

	travelers
	sum	mean	std
date
2019-03-31	189281658.0	2.103130e+06	282239.618354
2019-06-30	221756667.0	2.436886e+06	212600.697665
2019-09-30	220819236.0	2.400209e+06	260140.242892
2019-12-31	211103512.0	2.294603e+06	260510.040655
2020-03-31	155354148.0	1.726157e+06	685094.277420
2020-06-30	25049083.0	2.752646e+05	170127.402046
2020-09-30	63937115.0	6.949686e+05	103864.705739
2020-12-31	77541248.0	8.428397e+05	170245.484185
2021-03-31	86094635.0	9.566071e+05	280399.809061

Window calculations

Window calculations are similar to group by calculations except the group over which the calculation is performed isn't static – it can move or expand. Pandas provides functionality for constructing a variety of windows, including moving/rolling windows, expanding windows (e.g., cumulative sum or mean up to the current date in a time series), and exponentially weighted moving windows (to weight closer observations more than further ones). We will only look at rolling and expanding calculations here.

Source

Performing a window calculation is very similar to a group by calculation – we first define the window, and then we specify the aggregation:

tsa_melted_holiday_travel.loc['2020'].assign(
    **{
        '7D MA': lambda x: x.rolling('7D').travelers.mean(),
        'YTD mean': lambda x: x.expanding().travelers.mean()
      }
).head(10)

	year	travelers	holiday	7D MA	YTD mean
date
2020-01-01	2020	2311732.0	New Year's Day	2.311732e+06	2.311732e+06
2020-01-02	2020	2178656.0	New Year's Day	2.245194e+06	2.245194e+06
2020-01-03	2020	2422272.0	NaN	2.304220e+06	2.304220e+06
2020-01-04	2020	2210542.0	NaN	2.280800e+06	2.280800e+06
2020-01-05	2020	1806480.0	NaN	2.185936e+06	2.185936e+06
2020-01-06	2020	1815040.0	NaN	2.124120e+06	2.124120e+06
2020-01-07	2020	2034472.0	NaN	2.111313e+06	2.111313e+06
2020-01-08	2020	2072543.0	NaN	2.077144e+06	2.106467e+06
2020-01-09	2020	1687974.0	NaN	2.007046e+06	2.059968e+06
2020-01-10	2020	2183734.0	NaN	1.972969e+06	2.072344e+06

To understand what's happening, it's best to visualize the original data and the result, so here's a sneak peek of plotting with pandas. First, some setup to embed SVG-format plots in the notebook:

import matplotlib_inline
from utils import mpl_svg_config

matplotlib_inline.backend_inline.set_matplotlib_formats(
    'svg', # output images using SVG format
    **mpl_svg_config('section-2') # optional: configure metadata
)

Tip: For most use cases, only the first argument is necessary – we will discuss the second argument in more detail in the next section.

Now, we call the plot() method to visualize the data:

_ = tsa_melted_holiday_travel.loc['2020'].assign(
    **{
        '7D MA': lambda x: x.rolling('7D').travelers.mean(),
        'YTD mean': lambda x: x.expanding().travelers.mean()
      }
).plot(title='2020 TSA Traveler Throughput', ylabel='travelers', alpha=0.8)

Other types of windows:

• exponentially weighted moving: use the ewm() method
• custom: create a subclass of pandas.api.indexers.BaseIndexer or use a pre-built one in pandas.api.indexers

Exercise 2.5

Using the taxi trip data in the 2019_Yellow_Taxi_Trip_Data.csv file, resample the data to an hourly frequency based on the dropoff time. Calculate the total trip_distance, fare_amount, tolls_amount, and tip_amount, then find the 5 hours with the most tips.

Solution

import pandas as pd

taxis = pd.read_csv(
    '../data/2019_Yellow_Taxi_Trip_Data.csv',
    parse_dates=True, index_col='tpep_dropoff_datetime'
)
taxis.resample('1h')[[
    'trip_distance', 'fare_amount', 'tolls_amount', 'tip_amount'
]].sum().nlargest(5, 'tip_amount')

	trip_distance	fare_amount	tolls_amount	tip_amount
tpep_dropoff_datetime
2019-10-23 16:00:00	10676.95	67797.76	699.04	12228.64
2019-10-23 17:00:00	16052.83	70131.91	4044.04	12044.03
2019-10-23 18:00:00	3104.56	11565.56	1454.67	1907.64
2019-10-23 15:00:00	14.34	213.50	0.00	51.75
2019-10-23 19:00:00	98.59	268.00	24.48	25.74

Section 2 Complete 🎉

Section 3: Data Visualization

The human brain excels at finding patterns in visual representations of the data; so in this section, we will learn how to visualize data using pandas along with the Matplotlib and Seaborn libraries for additional features. We will create a variety of visualizations that will help us better understand our data.

Source

Learning Path

1. Why is data visualization necessary?
2. Plotting with pandas
3. Plotting with Seaborn
4. Customizing plots with Matplotlib

Why is data visualization necessary?

So far, we have focused a lot on summarizing the data using statistics. However, summary statistics are not enough to understand the distribution – there are many possible distributions for a given set of summary statistics. Data visualization is necessary to truly understand the distribution:

A set of points forming a panda can also form a star without any significant changes to the summary statistics displayed above. (source: Data Morph)

Learning Path

1. Why is data visualization necessary?
2. Plotting with pandas
3. Plotting with Seaborn
4. Customizing plots with Matplotlib

Plotting with pandas

We can create a variety of visualizations using the plot() method. In this section, we will take a brief tour of some of this functionality, which under the hood uses Matplotlib.

Once again, we will be working with the TSA traveler throughput data that we cleaned up in the previous section:

import pandas as pd

tsa_melted_holiday_travel = pd.read_csv(
    '../data/tsa_melted_holiday_travel.csv',
    parse_dates=True, index_col='date'
)
tsa_melted_holiday_travel.head()

	year	travelers	holiday
date
2019-01-01	2019	2126398.0	New Year's Day
2019-01-02	2019	2345103.0	New Year's Day
2019-01-03	2019	2202111.0	NaN
2019-01-04	2019	2150571.0	NaN
2019-01-05	2019	1975947.0	NaN

To embed SVG-format plots in the notebook, we will configure the Matplotlib plotting backend to generate SVG output (first argument) with custom metadata (second argument):

import matplotlib_inline
from utils import mpl_svg_config

matplotlib_inline.backend_inline.set_matplotlib_formats(
    'svg', # output images using SVG format
    **mpl_svg_config('section-3') # optional: configure metadata
)

Note: The second argument is optional and is used here to make the SVG output reproducible by setting the hashsalt along with some metadata, which will be used by Matplotlib when generating any SVG output (see the utils.py file for more details). Without this argument, different runs of the same plotting code will generate plots that are visually identical, but differ at the HTML level due to different IDs, metadata, etc.

Line plots

Let's continue with the example of rolling and expanding calculations:

plot_data = tsa_melted_holiday_travel.drop(columns='year').loc['2020'].assign(
    **{
        '7D MA': lambda x: x.travelers.rolling('7D').mean(),
        'YTD mean': lambda x: x.travelers.expanding().mean()
      }
)
plot_data.head()

	travelers	holiday	7D MA	YTD mean
date
2020-01-01	2311732.0	New Year's Day	2311732.0	2311732.0
2020-01-02	2178656.0	New Year's Day	2245194.0	2245194.0
2020-01-03	2422272.0	NaN	2304220.0	2304220.0
2020-01-04	2210542.0	NaN	2280800.5	2280800.5
2020-01-05	1806480.0	NaN	2185936.4	2185936.4

The plot() method will generate line plots for all numeric columns by default:

plot_data.plot(
    title='2020 TSA Traveler Throughput', ylabel='travelers', alpha=0.8
)

<Axes: title={'center': '2020 TSA Traveler Throughput'}, xlabel='date', ylabel='travelers'>

The plot() method returns an Axes object that can be modified further (e.g., to add reference lines, annotations, labels, etc.). Let's walk through an example.

Bar plots

For our next example, we will plot vertical bars to compare monthly TSA traveler throughput across years. Let's start by creating a pivot table with the information we need:

plot_data = tsa_melted_holiday_travel['2019':'2021-04']\
    .assign(month=lambda x: x.index.month)\
    .pivot_table(index='month', columns='year', values='travelers', aggfunc='sum')
plot_data.head()

year	2019	2020	2021
month
1	59405722.0	61930286.0	23598230.0
2	57345684.0	60428859.0	24446345.0
3	72530252.0	32995003.0	38050060.0
4	70518994.0	3322548.0	41826159.0
5	74617773.0	7244733.0	NaN

Pandas offers other plot types via the kind parameter, so we specify kind='bar' when calling the plot() method. Then, we further format the visualization using the Axes object returned by the plot() method:

import calendar
from matplotlib import ticker

ax = plot_data.plot(
    kind='bar', rot=0, xlabel='', ylabel='travelers',
    figsize=(8, 1.5), title='TSA Monthly Traveler Throughput'
)

# use month abbreviations for the ticks on the x-axis
ax.set_xticklabels(calendar.month_abbr[1:])

# show y-axis labels in millions instead of scientific notation
ax.yaxis.set_major_formatter(ticker.EngFormatter())

# customize the legend
ax.legend(title='', loc='center', bbox_to_anchor=(0.5, -0.3), ncols=3, frameon=False)

<matplotlib.legend.Legend at 0x103e10770>

Some additional things to keep in mind:

• Matplotlib's ticker module provides functionality for customizing both the tick labels and locations – check out the documentation for more information.
• Pandas supports horizontal and stacked bars as well; this article shows how to make stacked horizontal bars using a pivot table.
• The plot() method takes a lot of parameters, many of which get passed down to Matplotlib; however, sometimes we need to use Matplotlib calls directly.

Plotting distributions

Let's now compare the distribution of daily TSA traveler throughput across years. We will create a subplot for each year with both a histogram and a kernel density estimate (KDE) of the distribution. Pandas has generated the Figure and Axes objects for both examples so far, but we can build custom layouts by creating them ourselves with Matplotlib using the plt.subplots() function. First, we will need to import the pyplot module:

import matplotlib.pyplot as plt

While pandas lets us specify that we want subplots and their layout (with the subplots and layout parameters, respectively), using Matplotlib to create the subplots directly gives us additional flexibility:

# define the subplot layout
fig, axes = plt.subplots(3, 1, sharex=True, sharey=True, figsize=(6, 4))

for year, ax in zip(tsa_melted_holiday_travel.year.unique(), axes):
    plot_data = tsa_melted_holiday_travel.loc[str(year)].travelers
    plot_data.plot(kind='hist', legend=False, density=True, alpha=0.8, ax=ax)
    plot_data.plot(kind='kde', legend=False, color='blue', ax=ax)
    ax.set(title=f'{year} TSA Traveler Throughput', xlabel='travelers')

fig.tight_layout() # handle overlaps

Tip: If you're new to the zip() function, check out this article.

Exercise 3.1

Using the TSA traveler throughput data in the tsa_melted_holiday_travel.csv file, create box plots for traveler throughput for each year in the data. Hint: Pass kind='box' into the plot() method to generate box plots.

Solution

We start by handling imports:

from matplotlib import ticker
import pandas as pd

Next, we write a function that will read in the data, use the pivot() method to reshape it, and then use the plot() method to generate the box plot:

def ex1():
    df = pd.read_csv('../data/tsa_melted_holiday_travel.csv')
    plot_data = df.pivot(columns='year', values='travelers')
    ax = plot_data.plot(kind='box')
    ax.set(xlabel='year', ylabel='travelers', title='TSA Traveler Throughput')
    ax.yaxis.set_major_formatter(ticker.EngFormatter())
    return ax

Finally, we call our function:

ex1()

<Axes: title={'center': 'TSA Traveler Throughput'}, xlabel='year', ylabel='travelers'>

Learning Path

1. Why is data visualization necessary?
2. Plotting with pandas
3. Plotting with Seaborn
4. Customizing plots with Matplotlib

Plotting with Seaborn

The Seaborn library provides the means to easily visualize long-format data without first pivoting it. In addition, it also offers some additional plot types – once again building on top of Matplotlib. Here, we will look at a few examples of visualizations we can create with Seaborn.

Visualizing long-format data

With Seaborn, we can specify plot colors according to values of a column with the hue parameter. When working with functions that generate subplots, we can also specify how to split the subplots by values of a long-format column with the col and row parameters. Here, we revisit the comparison of the distribution of TSA traveler throughput across years:

import seaborn as sns

sns.displot(
    data=tsa_melted_holiday_travel, x='travelers', col='year', kde=True, height=2.5
)

<seaborn.axisgrid.FacetGrid at 0x13f538560>

Heatmaps

We can also use Seaborn to visualize pivot tables as heatmaps:

data = tsa_melted_holiday_travel['2019':'2021-04']\
    .assign(month=lambda x: x.index.month)\
    .pivot_table(index='month', columns='year', values='travelers', aggfunc='sum')

data

year	2019	2020	2021
month
1	59405722.0	61930286.0	23598230.0
2	57345684.0	60428859.0	24446345.0
3	72530252.0	32995003.0	38050060.0
4	70518994.0	3322548.0	41826159.0
5	74617773.0	7244733.0	NaN
6	76619900.0	14481802.0	NaN
7	79511968.0	20740781.0	NaN
8	74776010.0	21708071.0	NaN
9	66531258.0	21488263.0	NaN
10	72096495.0	25636496.0	NaN
11	68787654.0	25512987.0	NaN
12	70219363.0	26391765.0	NaN

ax = sns.heatmap(data=data / 1e6, cmap='Blues', annot=True, fmt='.1f')
_ = ax.set_yticklabels(calendar.month_abbr[1:], rotation=0)
_ = ax.set_title('Total TSA Traveler Throughput (in millions)')

Tip: Reference the Matplotlib documentation for more information on colormaps and named colors.

We're moving on from Seaborn now, but there is a lot more available in the API. Be sure to check out the following at a minimum:

• pairwise plots with pairplot()
• categorical scatter plots with swarmplot()
• joint distribution plots with jointplot()
• FacetGrids for custom layouts with any plot type

Exercise 3.2

Using the TSA traveler throughput data in the tsa_melted_holiday_travel.csv file, create a heatmap that shows the 2019 TSA median traveler throughput by day of week and month.

Solution

We start by reading in the data and handling imports:

import calendar

from matplotlib import ticker
import pandas as pd
import seaborn as sns

Next, we write a function that will read in the data, create a pivot table, and visualize it as a heatmap:

def ex2():
    df = pd.read_csv(
        '../data/tsa_melted_holiday_travel.csv',
        parse_dates=True, index_col='date'
    )
    plot_data = df.loc['2019'].assign(
        day_of_week=lambda x: x.index.dayofweek, month=lambda x: x.index.month
    ).pivot_table(
        index='day_of_week', columns='month', values='travelers', aggfunc='median'
    )

    ax = sns.heatmap(data=plot_data / 1e6, annot=True, fmt='.1f', cmap='Blues')
    ax.set_xticklabels(calendar.month_abbr[1:])
    ax.set_yticklabels(calendar.day_abbr, rotation=0)
    ax.set_title('2019 TSA Median Traveler Throughput\n(in millions)')

    return ax

Finally, we call our function:

ex2()

<Axes: title={'center': '2019 TSA Median Traveler Throughput\n(in millions)'}, xlabel='month', ylabel='day_of_week'>

Learning Path

1. Why is data visualization necessary?
2. Plotting with pandas
3. Plotting with Seaborn
4. Customizing plots with Matplotlib

Customizing plots with Matplotlib

In this final section, we will discuss how to use Matplotlib to customize plots. Since there is a lot of functionality available, we will only be covering how to add shaded regions and annotations here, but be sure to check out the documentation for more.

Adding shaded regions

When looking at a plot of TSA traveler throughput over time, it's helpful to indicate periods during which there was holiday travel. We can do so with the axvspan() method:

plot_data = tsa_melted_holiday_travel['2019-05':'2019-11']
ax = plot_data.travelers.plot(
    title='TSA Traveler Throughput', ylabel='travelers', figsize=(9, 2)
)
ax.yaxis.set_major_formatter(ticker.EngFormatter())

# collect the holiday ranges (start and end dates)
holiday_ranges = plot_data.dropna().reset_index()\
    .groupby('holiday').agg({'date': ['min', 'max']})

# create shaded regions for each holiday in the plot
for start_date, end_date in holiday_ranges.to_numpy():
    ax.axvspan(start_date, end_date, color='gray', alpha=0.2)

Tip: Use axhspan() for horizontally shaded regions and axvline() / axhline() for vertical/horizontal reference lines.

Adding annotations

We can use the annotate() method to add annotations to the plot. Here, we point out the day in 2019 with the highest TSA traveler throughput, which was the day after Thanksgiving:

plot_data = tsa_melted_holiday_travel.loc['2019']
ax = plot_data.travelers.plot(
    title='TSA Traveler Throughput', ylabel='travelers', figsize=(9, 2)
)
ax.yaxis.set_major_formatter(ticker.EngFormatter())

# highest throughput
max_throughput_date = plot_data.travelers.idxmax()
max_throughput = plot_data.travelers.max()
_ = ax.annotate(
    f'{max_throughput_date:%b %d}\n({max_throughput / 1e6:.2f} M)',
    xy=(max_throughput_date, max_throughput),
    xytext=(max_throughput_date - pd.Timedelta(days=25), max_throughput * 0.92),
    arrowprops={'arrowstyle': '->'}, ha='center'
)

Some things to keep in mind:

• We used Axes methods to customize our plots (i.e., an object-oriented approach), but the pyplot module provides equivalent functions (i.e., a functional approach) for adding shaded regions, reference lines, annotations, etc. – although the function names might be slightly different than their Axes method counterparts (e.g., Axes.set_xlabel() vs. plt.xlabel()).
• In general, try to stick to either the object-oriented or functional approach rather than mixing the two. However, be careful when working with subplots – pyplot functions will only affect the last subplot.
• The anatomy of a figure diagram in the Matplotlib documentation is a great resource for identifying which objects you will need to access for plot customizations.

For more on data visualization in Python, including animations and interactive plots, check out my Beyond the Basics: Data Visualization in Python workshop.

Exercise 3.3

Annotate the medians in the box plot from Exercise 3.1. Hint: The x coordinates will be 1, 2, and 3 for 2019, 2020, and 2021, respectively. Alternatively, to avoid hardcoding values, you can use the Axes.get_xticklabels() method, in which case you should look at the documentation for the Text class.

Solution

First, we modify the return statement from the solution to Exercise 3.1 to also give us the data:

from matplotlib import ticker
import pandas as pd


def ex1():
    df = pd.read_csv('../data/tsa_melted_holiday_travel.csv')
    plot_data = df.pivot(columns='year', values='travelers')
    ax = plot_data.plot(kind='box')
    ax.set(xlabel='year', ylabel='travelers', title='TSA Traveler Throughput')
    ax.yaxis.set_major_formatter(ticker.EngFormatter())
    return ax, plot_data

Now, we can build upon ex1() in a new function:

def ex3():
    ax, plot_data = ex1()

    # add annotations
    medians = plot_data.median()
    for tick_label in ax.get_xticklabels():
        median = medians[int(tick_label.get_text())]
        ax.annotate(
            f'{median / 1e6:.1f} M',
            xy=(tick_label.get_position()[0], median),
            ha='center', va='bottom'
        )

    return ax

Calling our function returns an annotated version of the plot from Exercise 3.1:

ex3()

<Axes: title={'center': 'TSA Traveler Throughput'}, xlabel='year', ylabel='travelers'>

Section 3 Complete 🎉

Section 4: Hands-On Data Analysis Lab

We will practice all that you’ve learned in a hands-on lab. This section features a set of analysis tasks that provide opportunities to apply the material from the previous sections.

Lab Tasks

This Google form contains all the tasks for this lab. You can submit as many times as you wish, and you will be able to see which questions you got wrong (if any).

Important notes:

• You don't have to answer all the questions to submit.
• There are likely more questions than you will have time to answer, so be sure to submit before the session ends.

Section 4 Complete 🎉

Related content

All examples herein were developed exclusively for this workshop – check out Hands-On Data Analysis with Pandas and my Beyond the Basics: Data Visualization in Python workshop for more.

stefaniemolin.com

Thank you!

I hope you enjoyed the session. You can follow my work on the following platforms:

stefaniemolin.com

github.com/stefmolin

twitter.com/StefanieMolin

linkedin.com/in/stefanie-molin