Pandas DataFrame | clip method
Start your free 7-days trial now!
Pandas DataFrame.clip(~) method is used to ensure that the values of the DataFrame are between a particular range.
Parameters
1. lowerlink | scalar or array_like | optional
The lower bound to clip. Values that are smaller than lower will be replaced by lower. You can ignore the lower bound by passing in None.
2. upperlink | scalar or array_like | optional
The upper bound to clip. Values that are larger than upper will be replaced by upper. You can ignore the upper bound by passing in None.
3. axislink | int or string | optional
Whether to perform clipping to each row or each column:
Axis | Description |
|---|---|
| Clip entire DataFrame. |
| Clip element-wise for each column. |
| Clip element-wise for each row. |
By default, axis=None. When you pass in an array for either lower or upper, you should set axis to either 0 or 1 depending on your needs. See examples below for clarification.
4. inplace | boolean | optional
Whether or not to directly modify the source DataFrame without creating a new one. By default, inplace=False, that is, a new DataFrame will be returned and the source DataFrame will be kept intact.
Return Value
A DataFrame with its values clipped.
Examples
Basic usage
Consider the following DataFrame:
df
A B0 2 61 4 8
Clipping with lower and upper bounds set to 3 and 6 respectively:
df.clip(lower=3, upper=6)
A B0 3 61 4 6
Here, note the following:
any value in the DataFrame that is smaller than
3will be replaced by the value3.any value that is larger than
6will be replaced by the value6.at the end, the lowest value in the DataFrame should be
3, and the largest should be6.
Passing in an array
Clipping each column
Consider the same DataFrame as above:
df
A B0 2 61 4 8
We can pass in an array of bounds, like so:
df.clip(lower=[5,2], axis=0)
A B0 5 61 4 8
Here, since axis=0, we are clipping each column element-wise. To break down what's happening:
Column A: [5 2] lower clips [2 4] = [5 4]Column B: [5 2] lower clips [6 8] = [6 8]
We could also set a combination of an array and a number like so:
df.clip(lower=[5,2], upper=6, axis=0)
A B0 5 61 4 6
To break down what's happening:
Column A: [5 2] lower clips [2 4] = [5 4] -> upper clipped to [5 4]Column B: [5 2] lower clips [6 8] = [6 8] -> upper clipped to [6 6]
Clipping each row
Here's df again for your reference:
df
A B0 2 61 4 8
Setting axis=1 clips each row element-wise:
df.clip(lower=[5,2], upper=6, axis=1)
A B0 5 61 5 6
Again, here's the break down:
Column One: [5 2] lower clips [2 6] = [5 6] -> upper clipped to [5 6]Column Two: [5 2] lower clips [4 8] = [5 8] -> upper clipped to [5 6]
