How Do I Melt and Manipulate Pandas DataFrames Using the `melt()` Function?

How do I melt a pandas dataframe?

Melting a pandas dataframe involves converting it from a wide format with multiple columns to a long format with two primary columns: 'variable' and 'value'. This can be achieved using the melt() function.

Problem 1: Melting a dataframe to transpose columns into rows

df = df.melt(id_vars=['Name', 'Age'], var_name='Subject', value_name='Grades')

This will convert the original dataframe into the desired format, with columns for 'Name', 'Age', 'Subject', and 'Grades'. The 'Subject' column will contain the original column headers, while 'Grades' will contain the corresponding values.

Problem 2: Filtering melted data to include specific columns

df = df.melt(id_vars=['Name', 'Age'], var_name='Subject', value_name='Grades')
filtered_df = df[df['Subject'] == 'Math']

By filtering the melted dataframe, you can remove unwanted columns. In this case, only the rows with 'Subject' set to 'Math' are retained.

Problem 3: Grouping melted data and ordering by scores

df = df.melt(id_vars=['Name', 'Age'], var_name='Subject', value_name='Grades')
sorted_df = df.groupby('value').agg({'Name': ', '.join, 'Subject': ', '.join})
sorted_df = sorted_df.sort_values('value')

This merges rows with duplicate values in the 'value' column, causing the 'Name' and 'Subject' columns to aggregate with comma-separated values. The result is then sorted to order the grades ascendingly.

Problem 4: Unmelting a melted dataframe

unmelted_df = df.pivot(index=['Name', 'Age'], columns='Subject', values='Grades')

To convert the melted dataframe back to its original format, you can use the pivot() function. This will group the data by the 'Name' and 'Age' columns and pivot the 'Subject' column to create the original dataframe shape.

Problem 5: Grouping melted data and concatenating subjects and grades

df = df.melt(id_vars=['Name', 'Age'], var_name='Subject', value_name='Grades')
grouped_df = df.groupby('Name').agg({'Subject': ', '.join, 'Grades': ', '.join})

Similar to Problem 3, this groups the melted dataframe by 'Name' and concatenates the 'Subject' and 'Grades' columns using commas. The result is a dataframe with one row per person, listing their subjects and grades.

Problem 6: Melting a dataframe with all columns as values

df = df.melt(value_name='Value', var_name='Column')

To melt the dataframe with all columns as values, omit the id_vars argument from the melt() function. This will create a dataframe where the 'Column' column lists the original column headers, and the 'Value' column contains the corresponding values.

The above is the detailed content of How Do I Melt and Manipulate Pandas DataFrames Using the `melt()` Function?. For more information, please follow other related articles on the PHP Chinese website!