How to deal with missing values in Azure Machine Learning Studio

Question

Looks like I have 672 mission values, according to statistics. There are NULL value in QuotedPremium column.

I implemented Clean Missing Data module where it should substitute missing values with 0, but for some reason I'm still seeing NULL values as QuotedPremium, but...it says that missing values are = 0

Here you see it tells me that missing values = 0, but there are still NULLs

So what really happened after I ran Clean Missing Data module? Why it ran succesfully but there are still NULL values, even though it tells that number of missing values are 0.

`NULL` is a value, hence such entries are neither cleaned nor reported as missing — desertnaut, Jan 04 '18 at 01:31
Oh yeah. You are right. I just expected all nulls become 0's. — Serdia, Jan 04 '18 at 01:38

score 2 · Accepted Answer · answered Jan 04 '18 at 01:43

2

NULL is indeed a value; entries containing NULLs are not missing, hence they are neither cleaned with the 'Clean Missing Data' operator nor reported as missing.

answered Jan 04 '18 at 01:43

desertnaut

57,590
26
140
166

score 1 · Answer 2 · answered Jan 17 '20 at 13:12

Since they are not really missing values, its a string NULL which is added to all these cells. So, in order to substitute these values with 0 you can use this below:

Use Execute R Script module, and add this code in it.

dataset1 <- maml.mapInputPort(1); # class: data.frame
dataset1[dataset1 == "NULL"] = 0; # Wherever cell's value is "NULL", replace it with 0
maml.mapOutputPort("dataset1"); # return the modified data.frame

Image for same:

How to deal with missing values in Azure Machine Learning Studio

2 Answers2