The** **pandas software library is written for Python and is mostly used for data analysis. It works as a data manipulation module.

The **pandas DataFrame **is a two-dimensional tabular data structure in which data is aligned in a tabular form in rows and columns.

The pandas Dataframe consists of three principal components:

- Principal Data
- Rows (placed left to right horizontally)
- Columns (placed top to bottom vertically)

**DataFrame.sub() **

Here, `sub()`

means subtraction, and this method performs subtraction operations on data frames. It is an element-wise operation and works like a binary subtraction ( **- **) operator.

DataFrame.sub (other, axis = 'columns', level = None, fill_value = None)

It has the following argument values:

`other`

: This parameter is a single or multiple element data structure or list-like object. It can be a DataFrame, series, sequence, scalar, or a constant.`axis`

: This is used for deciding the axis on which the operation is applied. Whether to compare by the index (0 or ‘index’) or columns (1 or ‘columns’) that is {0 or ‘index’, 1 or ‘columns’}.`level`

: This parameter is used to broadcast across a level and matching index values on the passed MultiIndex level. It is either a number or a label that indicates where to compare.`fill_value`

: This parameter is a number or None. It specifies what to do with NaN values before subtracting. If data in both corresponding DataFrame locations is missing, the result will be missing.

Here, the first parameter is required and the other three are optional**.**

This method returns a `dataFrame`

in the result** **obtained by subtraction of two DataFrames**, **or** **a DataFrame division with a scaler.

The first thing for implementation is to import pandas. Here we are importing pandas as `pd.`

So, `pd`

will be used in place of panda in the entire program.

# importing pandas as pdimport pandas as pd# Creating a dataframe with four observationsdf= pd.DataFrame({"ClassA":[100,50,10],"ClassB":[50,20,30],"ClassC":[70,70,25],"ClassD":[150,300,0]})# Print the dataframeprint(df)print()#subtractin of 10 from each and every valueprint(df.sub(10))

- Lines 4–7: We create a DataFrame including dictionaries having classes as keys.
- Line 9: We print the DataFrame.
- Line 12: Here the multiplication method is used that is
`df.sub(`

`10)`

When a single parameter (`10`

) is passed, it will be subtracted from every entry of the DataFrame.

# Subtract these elements from the respective classprint(df.sub([20, 10, 5, 1], axis='columns'))

We are subtracting 20 from the first class, 10 from the second class, 5 from the third class, and 1 from the fourth class where the axis is columns.

# importing pandas as pdimport pandas as pd# Creating a dataframe with three observationsdf= pd.DataFrame({"ClassA":[100,50,10],"ClassB":[50,20,30],"ClassC":[70,70,25],})# Print the dataframeprint(df)print()# subtracting with series type dataprint(df.sub(pd.Series([5, 10, 2], index=[0,1,2]), axis='index'))

Here, we have three elements in series `5`

, `10`

, and `20`

and indexes as `0`

, `1`

, and `2`

. Since the method is applied index wise `axis = index`

, the result is obtained in a way that the first series element `5`

will be subtracted from each value of the first index which is `0`

. The next series element `10`

, will be subtracted from each value of index `1`

and so on.

We can perform a variety of subtractions on one or more DataFrames just by changing parameters in different ways by using the `DataFrame.sub()`

method. In the case of any `fill_value`

parameter and assign it by the value we want written in place of the empty or missing values in the data instead of NaN.

TRENDING TOPICS