What is Pandas?

Pandas is an add-on library to Python.

It let’s us do more things with our code, specifically with dataframes.

Importing pandas

import pandas as pd

To analyze dataframes and load these csv files, we need to make sure that we bring in the pandas library into Python.

Before we start writing any valuable code for loading data and doing data analysis we need to import it with the following code.

Reading in Data

candy = pd.read_csv('data/candybars.csv')

candy

	name	weight	chocolate	peanuts	caramel	available_canada_america
0	Coffee Crisp	50	1	0	0	Canada
1	Butterfinger	184	1	1	1	America
2	Skor	39	1	0	1	Both
...	...	...	...	...	...	...
22	Almond Joy	46	1	0	0	America
23	Oh Henry	51	1	1	1	Both
24	Cookies and Cream	43	0	0	0	Both

25 rows × 6 columns

Next we can bring in our data named candybars which is stored as a .csv file.

Let’s break this up:

• pd is the short form for pandas, which we are using to manipulate our dataframe.
  
• read_csv() is the tool that does the job and, in this case, it is reading in the csv file named candybars.csv.
  
• candy is The dataframe is now saved as an object called candy.

The dataframe is stored in an object named candy and we can inspect in by “calling” the object name.

In these slides we can differentiate between the code that we typed in with a light grey background and it’s output which has a dark grey background.

From this dataframe, we can see that there are 25 different candy bars and 6 columns.

candy.columns

Index(['name', 'weight', 'chocolate', 'peanuts', 'caramel', 'available_canada_america'], dtype='object')

candy.shape

(25, 6)

We can obtain the names of the columns using .columns syntax, and if we wanted to see the dimensions of the whole dataframe we could use .shape after the dataframe name.

Breaking up the code, we interpret this as:

“From our dataframe that we saved as candy, tell me the columns andshape”

candy.head(2)

	name	weight	chocolate	peanuts	caramel	available_canada_america
0	Coffee Crisp	50	1	0	0	Canada
1	Butterfinger	184	1	1	1	America

candy.head()

	name	weight	chocolate	peanuts	caramel	available_canada_america
0	Coffee Crisp	50	1	0	0	Canada
1	Butterfinger	184	1	1	1	America
2	Skor	39	1	0	1	Both
3	Smarties	45	1	0	0	Canada
4	Twix	58	1	0	1	Both

What if we don’t want to output the whole table when displaying it as dataframe?

This can be really useful when we have dataframes that have hundreds or thousands of rows long.

We can specify how many rows of the dataset to show with .head() syntax.

.head(2) will output the first 2 rows of the dataframe.

We can specify any number of rows within the parentheses or we can leave it empty which will default to the first 5 rows.

Functions/Methods and Attributes

Attributes

Take candy.shape as an example.

In this case, our dataframe candy is our object and .shape is the attribute describing it.

Functions

In the example of pd.read_csv(), this function does the action of reading in our data.

Something you may have noticed is that when we use pd.read_csv() we put our instructions within the parentheses, whereas, when we use .shape or .head() the object that we are operating on comes before our desired command.

In Python, we use functions, methods and attributes. These are special words in Python that take instructions (we call these arguments) and do something.

Attributes can be distinguished from methods and functions as they do not have parentheses.

They can be thought of as nouns or adjectives that describe an object.

Take candy.shape as an example.

In this case, our dataframe candy is our object and .shape is the attribute describing it.

Functions and methods have parentheses.
They can be thought of as verbs that complete an action.

In the example of pd.read_csv(), this function does the action of reading in our data.

This is going to be discussed in more detail later in the course but now, simply be aware of the way we write the different instructions.

Comments

# This line does not execute anything. 

candy.shape  # This will output the shape of the dataframe

While we write code, it’s often useful to annotate it or include information for humans that we do not want to executed.

The easiest way to do this is with a hash (#) symbol. This creates a single line comment and prevents anything written after it from being executed by Python.

We use comments frequently in the exercises to help you understand what to do and what our intentions are.

It’s good practice to use them to explain our code so if we or someone else wants to read it at a later date, it’s easier to understand.

Let’s apply what we learned!