Time series data missing some observation

yield_df.csv (1.5 MB)
I am doing my final year project in crop yield prediction so i studied about RNNs for time series forecasting but the given dataset has some missing crops/observations after some time. What should i do?

I don’t have an answer to your question. There aren’t any missing values in your attached dataset (please see below). Am I missing something?

If you look at Albania at 1993 it does not have Sorghum and at 1994 it does not have Rice, paddy and Sorghum.

Albania has Sorghum only in the closed interval [1990, 1992].
The dataset covers years [1990, 2013]. Albania doesn’t have data about Sorghum from 1993 onwards.

Again, sorry I don’t understand your domain well enough guide you. To improve the odds of a response from others, consider adding more details about your dataset.

Since this question is off topic from the courses taught by deeplearning.ai, consider posting your query on sites like stackoverflow as well.

Good luck with your project.

1 Like

Most likely you’re going to have to make up some data to fill in the holes. The mean of the adjacent values is a common method.


Based on screenshot of mentor balaji your missing values are for 28242 for 6 values which you can divide by creating a data frame once with the missing values and another without missing values, and then compare by plot.

if you do not see too much variation, then you can ignore the missing values, but in case there is too many variations, you need to find ways to replace those missing values.


Thank you so much you:heart:. I will try what you said.

1 Like