convert rows to columns in pandas

How to Convert Columns into Rows in Pandas

Pandas is a popular module that helps you to import, export, transform data in Python. It offers tons of features to play with data, stored as tables. Often you may need to transpose columns into rows and vice versa, in your data. In this article, we will learn how to convert columns into rows in Pandas.


How to Convert Columns into Rows in Pandas

Let us say you have the following data in Python pandas dataframe df.

>>> print df
location    name    Jan-2020    Feb-2020    March-2020
A           "test"  12          20          30
B           "foo"   18          20          25

Let us say you want to transform it into the following.

location    name    Date        Value
A           "test"  Jan-2020    12       
A           "test"  Feb-2020    20
A           "test"  March-2020  30
B           "foo"   Jan-2020    18       
B           "foo"   Feb-2020    20
B           "foo"   March-2020  25

To perform the above transformation, we use melt() function. Here is its syntax.

pandas.melt(frame, id_vars=None, value_vars=None, var_name=None, value_name='value', col_level=None, ignore_index=True)

In the above function, id_vars is the list of columns to be retained after transformation, var_name is the column to unpivot, and value_name is the value of the column to be used for unpivotting.

Here is the command to convert rows into columns, by unpivotting the separate date columns into a single column.

df.melt(id_vars=["location", "name"], 
        var_name="Date", 
        value_name="Value")

  location    name        Date  Value
0        A  "test"    Jan-2010     12
1        B   "foo"    Jan-2010     18
2        A  "test"    Feb-2010     20
3        B   "foo"    Feb-2010     20
4        A  "test"  March-2010     30
5        B   "foo"  March-2010     25

When you run the above command what python pandas does is to pivot down all columns in input dataframe except the ones listed as id_vars, that is, location & name. While doing so, it will take all the remaining column names, create a new column called ‘Date’ and store these column names as Date column’s values. It will also take all the column values of the above mentioned column, create another column called Value and store all these values in it.

In this article, we have learnt how to convert columns into rows in Pandas.

Also read:

How to Set/Unset Cookie with jQuery
How to Abort AJAX Request in JavaScript/jQuery
How to Get Value of Text Input Field With JavaScript
How to Merge Two Arrays in JavaScript
How to Persist Variables Between Page Loads

Leave a Reply

Your email address will not be published.