Methods for dealing with missing values in time-series data

When working with time series, we often find missing data; the missing value is caused by sensors that have been disconnected or a data reader or recorder that has failed.

Within pandas library, there are few methods available to solve the missing value.

Let’s import the required libraries

input data frame

The following code displays data frame which contain some missing value

isna().sum() returns the summary of missing value in each column

Handling missing value by replacing the new value

when using the ffill method, the previous value is used to fill in the NaN value. Backward fill, on the other hand, is done with bfill, which used the next value to fill in.

Lastly, the interpolation method is particularly useful for data that is presented in a sequential or time-series format.

This post has shown the very basic method of filling missing data by using ffill, bfill and interpolation that useful for time-series imputation.

Thank you for reading till the end.