CSE 163

Pandas�

Suh Young Choi�

🎶 Listening to: Death’s Door soundtrack

💬 Before Class: What is the best polygon?

​

​

This Time

• Imports
• Pandas!
• How to read a file
• Accessing columns
• Element-wise operators
• Filtering
• loc

Last Time

• Dictionary Methods
• How to loop over dictionary
• CSVs
• List of Dictionaries

2

Announcements

• Data Exploration + LR 1 due tonight!

​

• THA 1 due on Thursday!

3

Checking in with the Check-ins

Lost in translation?

• Pseudocode might help get ideas out before grammar
• For Java users, check out Java 2 Python 3 by Hunter Schafer!

​

​

Concerned about falling behind?

• Reaching out might be scary, but having no help could be scarier!
• Office hours, Ed discussion board (you can be private or anonymous!)
• In-class and section announcements to track due dates

4

What about grading and the final project?

5

Recap

Imports

• A way for us to get functionality from modules we didn’t write ourselves!

​

Pandas

• New data types– DataFrames and Series
• More tabular-looking than the list of dictionaries

6

Importing

Importing lets you use the contents defined in another Python file

• We call a Python file a module
• Generally there are three ways to import!

​

7

# Method 1: Import module

import module

module.function()

​

# Method 1: Import module

import module

module.function()

​

# Method 2: Import and rename module

import module as m

m.function()

​

# Method 1: Import module

import module

module.function()

​

# Method 2: Import and rename module

import module as m

m.function()

​

# Method 3: Import specific function from module

from module import function

function()

​

DataFrame

• One of the basic data types from pandas is a DataFrame
• It’s essentially a table with column and rows!

8

Columns

Index (row)

Series

• A Series is like a 1-dimensional DataFrame (no columns!)
• Has an index
• It’s basically like a fancy dictionary/list hybrid
• For example

​

9

df['name']

​

df['name'][1] # 'Burma'

​

​

Series Operations

• You can do Series-wide operations without loops!

​

​

​

​

​

​

​

​

​

​

10

df['name’] += “ :)”

​

df[‘magnitude’] *= 2

Series Operations

• A leading question: What does this operation turn into?

11

df['name’] == “Burma”

​

Series Operations

• A leading question: What does this operation turn into?

12

df['name’] == “Burma”

​

Series Operations

• A leading question: What does this operation turn into?

13

df['name’] == “Burma”

​

Filtering

• Can use a bool Series to select which rows from the dataset

​

​

​

​

​

​

�

​

• Can use multiple filters with: & (and), | (or), ~ (not)

​

​

14

mask = df['magnitude'] > 5

df[mask]

# Same as: data[data['magnitude'] > 5]

​

df[(df['magnitude'] > 5) & ~(df['day'] == 27)]

Location

How to access data in pandas

Series

​

​

DataFrame

​

​

​

Options for indexers:

• A single value
• A list of values or a slice
• A mask
• : to select all values

Remember the end of a slice is inclusive unlike Python’s standard

15

df[<indexer>]

df.loc[<row indexer>, <column indexer>]

series[<indexer>]

Before Next Time

• Complete Lesson 7
• Remember not for points, but do go towards Checkpoint Tokens
• Data Exploration + LR 1 due today!

Next Time

• More Pandas functions!
• Specifically, groupby and apply

​

16