Not able to remove missing values, error with Multi-layer perceptron

#1

Hello 

I am not able to fillter out missing values using 'missing values' node.. i tried other nodes also but none seem to work.

WARN  Missing Value        4:19       Column "hail" still contains missing values.
Column "severity" still contains missing values.
Column "seed-tmt" still contains missing values......................
Column "mycelium" still contains missing values.

I am getting below error:

ERROR MultiLayerPerceptron Predictor 4:9        Execute failed: Input DataTable should not contain missing values.

Please help me selove this issue.

 

0 Likes

#2

Hi,

did you attach the correct workflow_ it does not contain a missing value node. 

About the missing value node, which options did you use to resolve the missings? Can you post a screenshot of the configuration dialog?

Best, Iris 

1 Like

#3

I have the same problem. Here is a screenshot of my workflow. What should I do to fix this error "ERROR MultiLayerPerceptron Predictor 0:18 Execute failed values?

1 Like

#4

Hi @Freud

Welcome to the KNIME Community :slight_smile:

It seems that your data has some missing values. Have you already tried the Missing Value node? With it, you can correct or remove the missing values. By doing it before the model, you will not get the error message.

Hope it helps!

Cheers,
Ana

1 Like

#5

@ana_ved Can you please tell me how to remove missing values? I only see a fix.

0 Likes

#6

Which strategies should be used to remove missing values?

0 Likes

#7

Hi Freud,

If you scroll down on this window, you will see an option named “remove row*”

image

0 Likes

#8

@ana_ved Do I understand correctly that from such a table, after selecting “remove row”, only the data from the third row will remain, and the rest will be deleted?rows

2 Likes

#9

That is true - but in this case this will happen because you are selecting to perform this operation to all columns that are number. In case you want to remove rows based on missing values in only one column, you can go to the tab “column settings” and pick a column, like in this example:

.

Bringing to your example: if you selected the first column, for instance, only the first and last row would be removed.

0 Likes

#10

@ana_ved Is it possible not to delete rows? Just so that the value of the missing value does not go further on?

0 Likes

#11

I mean like thatrows2

0 Likes

#12

Hi @Freud!

So, I am not sure if you can skip missing values. I think you can deal with them - being it removing the values or replacing them according to a certain rule. For instance, as you did before, you can replace it for a value you set, you can replace it by the average value of your data, and so on and so on.

So in the end your strategy to deal with missing values depends on your data and your use case.

Cheers,
Ana

2 Likes