Pandas Cheatsheet

ThePandas cheatsheet provides a fundamental reference to all the core concepts of pandas. This powerful library inPython is used for data manipulation, analysis, and handling structured data with ease. Whether you're working with large datasets, performing data cleaning, or analyzing trends, this cheat sheet will help you navigate Pandas easily. Go through the cheatsheet and learn thePython pandas library.

• Introduction to Pandas
• Installing Pandas
• Creating DataFrames
• Creating Series
• Reading Data
• Writing Data
• Selecting Columns
• Selecting Rows
• Filtering Data
• Boolean Indexing
• Querying Data
• Handling Missing Values
• Changing Data Types
• Renaming Columns
• Duplicates
• Replacing Values
• Sorting Data
• GroupBy
• Pivot Tables
• Apply Functions
• Merging and Joining
• Summary Statistics
• Value Counts
• Correlation
• Cumulative Functions
• MultiIndex
• Time Series Analysis
• Working with JSON
• Visualization

1. Introduction to Pandas

In theintroduction, Pandas is a popular open-source library in Python for data analysis. It provides data structures and functions to processes large datasets which includes tabular data such as spreadsheets and SQL tables. Here, we will learn how to import the pandas library.

import pandas as pd

2. Installing Pandas

Toinstall Pandas on the system, use the following command −

pip install pandas

3. Creating DataFrames

Thedataframe can be created using lists, dictionaries, and external data sources.

# Creating a DataFrame from a dictionaryimport pandas as pdinp_data = {"Name": ["Ravi", "Faran"], "Age": [25, 30]}df = pd.DataFrame(inp_data)print(df)

4. Creating Series

In Pandas, theseries is like a column in the table. You can create the pandas series using a list or NumPy arrays.

import pandas as pds = pd.Series([10, 20, 30, 40])

5. Reading Data

There are four methods toread data in Pandas − CSV, Excel, JSON, and SQL files.

# Reading a CSV filedf = pd.read_csv("data.csv")

6. Writing Data

Towrite the pandas dataframe in a CSV file, the user needs Dataframe.to_csv().

# Writing a DataFrame to a CSV filedf.to_csv("output.csv", index=False)

7. Selecting Columns

To select the specific column from dataframe −

# Selecting a single columndf["Name"]

8. Selecting Rows

To retrieve specific rows usingindex selection andslicing, Pandas provides the head() and tail() methods. The head() method returns the first few rows of the DataFrame, while the tail() method retrieves the last few rows.

df.head(5)

Or,

df.tail(5)

9. Filtering Data

Filtering data in pandas means it applies some conditions based on certain rows and columns.

# Filtering rows where Age > 25df[df["Age"] > 25]

10. Boolean Indexing

In pandas,boolean indexing means the process of filtering data using a boolean array.

mask = df["Age"] > 25df[mask]

11. Querying Data

In Pandas,querying data filters the dataframe by passing the condition as a string that returns matching rows. You can use the query() method.

df.query("Age > 25")

12. Handling Missing Values

To handle themissing values in Pandas, use the methods likedropna() andfillna(). Below is the implementation −

df.fillna(0, inplace=True)

Or,

import pandas as pd# Creating a DataFrame with missing valuesdata = {"Name": ["Vivek", "Faran", None, "Revathi"],        "Age": [25, None, 30, 35]}df = pd.DataFrame(data)# Dropping rows with missing valuesdf_result = df.dropna()print(df_result)

13. Changing Data Types

To convert thedata types in Python use the method astype(). This ensures the proper formatting.

df["Age"] = df["Age"].astype(int)

14. Renaming Columns

The easier way torename the columns in Pandas, use the method rename(). The following syntax is given below −

df.rename(columns={"old_name": "new_name"}, inplace=True)

15. Duplicates

To remove theduplicates from the rows, use the method drop_duplicates().

df.drop_duplicates(inplace=True)

16. Replacing Values

The term "replacing" is also known as "removing". To remove the specific values in a dataframe, use the methodreplace().

df["column_name"].replace({"old_value": "new_value"}, inplace=True)

17. Sorting Data

In Python, Pandas is a popular library that provides a built-in method calledsort_values(). This method allows users to sort the values of a DataFrame or Series in ascending or descending order.

import pandas as pddata = {'Name': ['Alex', 'John', 'Sunny', 'Usha'],        'Id': [2115, 6330, 8135, 4110],        'Score': [85, 90, 95, 80]}df = pd.DataFrame(data)# Sorting by 'Id' in ascending ordersorted_df = df.sort_values(by='Id')print(sorted_df)

18. GroupBy

GroupBy is used to split the data into groups based on some criteria and then apply a function to each group. Thus, this helps in data summarization and analyzing.

# Grouping by 'Gender' and calculating the mean agedf.groupby('Gender')['Age'].mean()

19. Pivot Tables

In Pandas, the use ofpivot tables is to summarize the data that allows users to aggregate data across multiple dimensions.

df.pivot_table(values='Age', index='Gender', columns='City', aggfunc='mean')

20. Apply Functions

In pandas, theapply() function is used to apply a function along the axis of a DataFrame or Series.

df.apply(lambda x: x.max() - x.min())

21. Merging and Joining

In Pandas, the concept ofmerging and joining allows users to combine multiple dataframes based on shared columns or indexes.

# Merging two DataFrames on a common column 'ID'df1.merge(df2, on='ID')

Or,

df1.join(df2, on='column_name', how='inner')

Explanation of join() parameters −

• on: Specifies the column or index to join on.
• how: This determine the type of join used for the dataset.

22. Summary Statistics

Summary statistics help in understanding the distribution and key properties of the dataset. Methods like 'mean()', 'median()', and 'std()' provide insights of the data from the given datasets.

# Getting summary statisticsdf.describe()

23. Value Counts

Thevalue_counts() method is used to get the frequency of unique values in a column.

df['col_name'].value_counts()

24. Correlation

Correlation means the relationship between two variables. The corr() method calculates the correlation coefficient between columns in a DataFrame.

df.corr()

25. Cumulative Functions

In Pandas,cumulative functions are those functions that add up or multiply values sequentially over time. You can use methods like cumsum() and cumprod().

df['Age'].cumsum()

26. MultiIndex

TheMultiIndex is a very simple concept that adds multiple levels of indexing in a DataFrame. So, it is possible to handle the complex data structure.

arrays = [['A', 'A', 'B', 'B'], [1, 2, 1, 2]]index = pd.MultiIndex.from_arrays(arrays, names=('Letter', 'Number'))df_multi = pd.DataFrame({'Value': [10, 20, 30, 40]}, index=index)

27. Time Series Analysis

Time series analysis works with time-indexed data. Pandas provides functionality to handle time series data by containing date parsing and resampling.

# Converting a column to datetime formatdf['Date'] = pd.to_datetime(df['Date'])# Resampling data by monthdf.resample('M').mean()

28. Working with JSON

JSON (JavaScript Object Notation) is a popular data format. In pandas, we have two ways to implement JSON −

• read_json() − It read the JSON data into a dataframe.
• to_json() − It convert dataframes into JSON format.

# read JSON data into a DataFramedf = pd.read_json('data.json')# convert DataFrame into JSONdf.to_json('output.json')

29. Visualization

Data visualization is key to understanding patterns and insights. Pandas integrates with libraries like Matplotlib and Seaborn to create various plots from DataFrames.

# Plotting a line graph using Pandasdf['Age'].plot(kind='line')# Plotting a histogramdf['Age'].plot(kind='hist', bins=10)