25+ Helpful Python Commands in Jupyter Notebook

Jupyter Notebook is an extremely helpful open-source web app used for all types of data transformation, modeling, and visualization. Below is a quick refresher on installing Jupyter Notebook and using Python commands to handle data sets within the app. Let's jump right in!

Create and activate a virtual environment

macOS Terminal

User-Macbook:~ user$ cd desktop

User-Macbook:desktop user$ python3 -m venv jup

User-Macbook:desktop user$ cd jup

User-Macbook:jup user$ source bin/activate

Windows Command Prompt

C:\Users\Owner> cd desktop

C:\Users\Owner\desktop> py -m venv jup

C:\Users\Owner\desktop> cd jup

C:\Users\Owner\desktop\jup> Scripts\activate

First, type python3 -m venv env for Mac or py -m venv env for Windows to create a virtual environment named jup. Then activate the virtual environment so you can properly install JupyterLab in the next step.

Install JupyterLab

macOS Terminal

(jup)User-Macbook:jup user$ pip install jupyterlab

Windows Command Prompt

(jup)C:\Users\Owner\desktop\jup> pip install jupyterlab

Install JupyterLab, the interactive development environment for notebooks, with the command pip install juypterlab. This may take a couple of minutes.

Open JupyterLab in the browser

macOS Terminal

(jup)User-Macbook:jup user$ jupyter notebook

Windows Command Prompt

(jup)C:\Users\Owner\desktop\jup> jupyter notebook

Once the installation is complete, open JupyterLab in your browser with the simple command juypter notebook.

Your virtual environment folders will then appear in the browser window. 

Create a new notebook

Click on the "New" button on the right side of the screen then select, "Python3". A notebook will then be created in a new tab.

Install numpy pandas nltk

import sys
!{sys.executable} -m pip install numpy pandas nltk

Type in the command pip install numpy pandas nltk in the first cell. Click Shift + Enter to run the cell's code. An asterisk will then appear in the brackets indicating it is running the code.

When finished, a new cell will appear below. You are now ready to use Python commands to load-in and clean your data.

Loading in your file

Load in a JSON file

import pandas as pd
df = pd.read_json(r'C:\Users\Owner\Desktop\data.json')

Load in a CSV file

import pandas as pd
df = pd.read_csv(r'C:\Users\Owner\Desktop\data.csv')

Load in an excel file

import pandas as pd
df = pd.read_excel(r'C:\Users\Owner\Desktop\data.xls')

In a new cell import pandas as pd. Then set a variable, in this case df for dataframe, as pd.read_doctype(r'full_path_to_file'). Run the cell.

Working with columns

display column names

df.columns

Lists the names of all of the columns in the data frame.

display specific columns

df['column_name']

df['column_name1', 'column_name2', 'column_name3']

Display one specific column or multiple columns by calling on their names.

sort alphabetically within columns

df.sort_values('column_name')

df.sort_values('column_name', ascending=False)

Display the data frame alphabetically by the column specified. Add ascending=False to sort by descending alphabetical order.

drop/delete columns

df = df.drop(columns = ['column_name']))

Drops the column specified from the data set.

create a new column with addition

df['new_column'] = df['column_name1'] + df['column_name2']

Creates a new column equal to the sum of column 1 plus column 2 data.

Working with rows

display first rows

df.head()

df.head(10)

Displays the first 5 rows of the data frame. Add a number in the parentheses and that number of rows will display.

display last rows

df.tail()

df.tail(10)

Displays the last 5 rows of the data frame. Add a number in the parentheses and that number of rows will display.

display a set of rows

df.iloc[0:5]

Displays row 0-4 of the data frame. Note, the data frame starts counting rows/columns at 0 instead of 1.

Working with the columns and rows

locate a specific value by row and column

df.iloc[2,1]

Displays a specific value in a designated location. The example above locates the value in row 2, column 1.

locate specific rows by column

df.loc[df['column_name'] == 'column_value']

Displays all of the rows that meet the specified column value.

locate specific rows that adhere to all of the column values

df.loc[(df['column_1'] == 'column_value') & df.loc(df['column_2'] == 'column_value')]

Displays all rows with the specific column value of column 1 and the specific column value of column 2.

locate specific rows that adhere to one of the column values

df.loc[(df['column_1'] == 'column_value') | df.loc(df['column_2'] == 'column_value')]

Displays all rows with the specific column value of column 1 or the specific column value of column 2.

locate specific rows in a column that contain a certain word

df.loc[df['column_1'].str.contains('word')]

Lists all of the rows in column 1 that contain the word specified.

drop specific rows in a column that contain a certain word

df.loc[~df['column_1'].str.contains('word')]

Drop all of the rows in column 1 that contain the word specified.

Changing data types

view data types

df.dtypes

Outputs the data type (i.e. object, int32, datetime64...) of each column.

change data to integer

df.column_name.astype(int)

Changes the data type of the column specified from a string to an integer.

change data to datetime field

pd.to_datatime(df.column_name)

Changes the data type of the column specified from a string to a date/time integer.

Working with the data set as a whole

reset the index

df.reset_index(drop=True, inplace=True)

Resets the index of the data frame. This works well if you delete or change the ordering of rows.

list duplicates

df.drop_duplicates()

Identifies duplicates in the data set.

drop duplicates

df.drop_duplicates()

Drops any duplicate rows from the data set.

combine datasets

data = pd.concat([df1, df2, df3])

Combines all of the data from each data frame into a new data set.

Saving updated data frames as files

JSON file

df.to_json('name_of_new_file.json')

Save the existing data frame as a new CSV file. 

CSV file

df.to_csv('name_of_new_file.csv')

Save the existing data frame as a new CSV file. 

excel file w/o the index

df.to_excel('name_of_new_file.xlsx', index=False)

Save the existing data frame as a new excel file without the index numbers. Note, index=False can be added to the other.