Python Pandas - Interpolation of Missing Values

---

---

Interpolation is a powerful technique in Pandas that used for handling the missing values in a dataset. This technique estimates the missing values based on other data points of the dataset. Pandas provides theinterpolate() method for both DataFrame and Series objects to fill in missing values using various interpolation methods.

In this tutorial, we will learn about theinterpolate() methods in Pandas for filling the missing values in a time series data, numeric data, and more using the different interpolation methods.

Basic Interpolation

The Pandasinterpolate() method of the both DataFrame and Series objects is used to fills the missing values using different Interpolation strategies. By default, Pandas automatically uses linear interpolation as the default method.

Example

Here is a basic example of calling theinterpolate() method for filling the missing values.

import numpy as npimport pandas as pddf = pd.DataFrame({"A": [1.1, np.nan, 3.5, np.nan, np.nan, np.nan, 6.2, 7.9],"B": [0.25, np.nan, np.nan, 4.7, 10, 14.7, 1.3, 9.2],})print("Original DataFrame:")print(df)# Using the  interpolate() methodresult = df.interpolate()print("\nResultant DataFrame after applying the interpolation:")print(result)

Following is the output of the above code −

Original DataFrame:

	A	B
0	1.1	0.25
1	NaN	NaN
2	3.5	NaN
3	NaN	4.70
4	NaN	10.00
5	NaN	14.70
6	6.2	1.30
7	7.9	9.20

Resultant DataFrame after applying the interpolation:

	A	B
0	1.100	0.250000
1	2.300	1.733333
2	3.500	3.216667
3	4.175	4.700000
4	4.850	10.000000
5	5.525	14.700000
6	6.200	1.300000
7	7.900	9.200000

Different Interpolating Methods

Pandas supports several interpolation methods, including linear, polynomial, pchip, akima, spline, and more. These methods provide flexibility for filling the missing values depending on the nature of your data.

Example

The following example demonstrates using theinterpolate() method with thebarycentric interpolation technique.

import numpy as npimport pandas as pddf = pd.DataFrame({"A": [1.1, np.nan, 3.5, np.nan, np.nan, np.nan, 6.2, 7.9],"B": [0.25, np.nan, np.nan, 4.7, 10, 14.7, 1.3, 9.2],})print("Original DataFrame:")print(df)# Applying the interpolate() with Barycentric methodresult = df.interpolate(method='barycentric')print("\nResultant DataFrame after applying the interpolation:")print(result)

Following is the output of the above code −

Original DataFrame:

i	A	B
0	1.1	0.25
1	NaN	NaN
2	3.5	NaN
3	NaN	4.70
4	NaN	10.00
5	NaN	14.70
6	6.2	1.30
7	7.9	9.20

Resultant DataFrame after applying the interpolation:

	A	B
0	1.100000	0.250000
1	2.596429	57.242857
2	3.500000	24.940476
3	4.061429	4.700000
4	4.531429	10.000000
5	5.160714	14.700000
6	6.200000	1.300000
7	7.900000	9.200000

Handling Limits in Interpolation

By default, Pandas interpolation fills all the missing values, but you can limit how many consecutive NaN values are filled using thelimit parameter of theinterpolate() method.

Example

The following example demonstrates filling the missing values of a Pandas DataFrame by limiting the consecutive fills using thelimit parameter of theinterpolate() method.

import numpy as npimport pandas as pddf = pd.DataFrame({"A": [1.1, np.nan, 3.5, np.nan, np.nan, np.nan, 6.2, 7.9],"B": [0.25, np.nan, np.nan, 4.7, 10, 14.7, 1.3, 9.2],})print("Original DataFrame:")print(df)# Applying the interpolate() with limitresult = df.interpolate(method='spline', order=2, limit=1)print("\nResultant DataFrame after applying the interpolation:")print(result)

Following is the output of the above code −

Original DataFrame:

i	A	B
0	1.1	0.25
1	NaN	NaN
2	3.5	NaN
3	NaN	4.70
4	NaN	10.00
5	NaN	14.70
6	6.2	1.30
7	7.9	9.20

Resultant DataFrame after applying the interpolation:

i	A	B
0	1.100000	0.250000
1	2.231383	-1.202052
2	3.500000	NaN
3	4.111529	4.700000
4	NaN	10.000000
5	NaN	14.700000
6	6.200000	1.300000
7	7.900000	9.200000

Interpolating Time Series Data

Interpolation can be applied to the Pandas time series data as well. It is useful when filling gaps in missing data points over time.

Example

Example statement −

import numpy as npimport pandas as pdindx = pd.date_range("2024-01-01", periods=10, freq="D")data = np.random.default_rng(2).integers(0, 10, 10).astype(np.float64)s = pd.Series(data, index=indx)s.iloc[[1, 2, 5, 6, 9]] = np.nanprint("Original Series:")print(s)result = s.interpolate(method="time")print("\nResultant Time Series after applying the interpolation:")print(result)

Following is the output of the above code −

Original Series:

Date	Value
2024-01-01	8.0
2024-01-02	NaN
2024-01-03	NaN
2024-01-04	2.0
2024-01-05	4.0
2024-01-06	NaN
2024-01-07	NaN
2024-01-08	0.0
2024-01-09	3.0
2024-01-10	NaN

Resultant Time Series after applying the interpolation:

Date	Value
2024-01-01	8.000000
2024-01-02	6.000000
2024-01-03	4.000000
2024-01-04	2.000000
2024-01-05	4.000000
2024-01-06	2.666667
2024-01-07	1.333333
2024-01-08	0.000000
2024-01-09	3.000000
2024-01-10	3.000000