"Impute Missing Value" example set is empty

pablo_199

im running a basic titanic data cleansing, and im trying to fill the na's of the cabin column with the impute missing value operator. i insert the missing values as input, i can even visualize them if i dragg to result, but once i connect it to Impute Missing Value, it says the example set is empty, could anyone help

varunm1

Hello @pablo_199

The issue in your process is that you are selecting a single attribute that has missing values in "impute missing values" operator. This doesn't work as the KNN needs predictor attributes to learn and impute missing values in a column. For this purpose, I selected all attributes. Once more issue is with filter examples before impute missing values, this will get only examples with missing value columns. The issue here is KNN doesn't have a clue what it needs to place in "Cabin" column as all the values in that column are missing due to this filter. I removed that.

Please find attached .rmp file. To run this go to File __> Import Process and point it to the file location you download into PC. Please let us know if you have any questions.

pablo_199

varunm1

Hello @pablo_199

Can you post your process by exporting using FILE --> Export process and attach the process here?

We will check it.

Thanks
Varun

pablo_199

sorry, the correct process is the one named so, ignore the previous one

pablo_199

Hello @varunm1
thanks for your reply, i have added the proccess now
thanks in advance for the help

varunm1

Hello @pablo_199

Thanks, Can you provide dataset as well? I thought you used titanic from Rapidminer samples, but its unable to retrieve.

I also see that the screenshot you mentioned earlier doesn't match with the current process you provided. The "impute missing values" operator is missing in the attached process.

lionelderkrikor

Hi @pablo_199, Hi @varunm1,

According the error "example set is empty", It seems that the Filter examples operator is filtering all the examples. (the condition missing attributes seems to be always true)
I think you don't need to add this operator for the task you want to aschieve ("trying to fill the na's of the cabin column").
Try to remove Filter Examples operator...
But I agree with Varun : Can you please share your data in order we can run your process ?

Thanks and regards,

Lionel

pablo_199

hi @varunm1 @lionelderkrikor
yes i submitted a wrong proccess and since i couldn`t erase it i added the right one. ill attach the dataset now, i tried both erasing the operator and adding it but the error remains the same. i want to thank you for the time in attempting to help me 

varunm1

Hello @pablo_199

The issue in your process is that you are selecting a single attribute that has missing values in "impute missing values" operator. This doesn't work as the KNN needs predictor attributes to learn and impute missing values in a column. For this purpose, I selected all attributes. Once more issue is with filter examples before impute missing values, this will get only examples with missing value columns. The issue here is KNN doesn't have a clue what it needs to place in "Cabin" column as all the values in that column are missing due to this filter. I removed that.

Please find attached .rmp file. To run this go to File __> Import Process and point it to the file location you download into PC. Please let us know if you have any questions.