I have a csv file and i am trying to replace with missing values.I am using 2 nodes.
One csv reader node and the other missing values node . There is 3 different value
1.Number ( double ) replace method MEAN
2.String replace method Most frequent Values
3.Number (Integer) replecae method MEDIAN
Even if i applied and executed this node there is still NA values.I can understand that the NA values are not the missing values.But how can i handle with them? Can your show me way to step by step ?
Can you describe the problem you have?
The Missing Value Node has an option to replace missings with the mean value and you can enter a
default String (“Fix Value”). So what is it you need that is not available?
Hello my friend
First of all i am sorry about i can not be more specific.Shortly there are NA values in my csv file .How can solve this problem ? I heard yesterday String Replace node but i can not use this.Can you help me more spesific ?
Have a good day
The Math Formula Node has a isNaN(x) function to detect those type of values.
You can use this in your Math Formula Node: if(isNaN($Value$)==1,"", $Value$) for your numerical columns. This will create Missing Values instead of NaNs that you can handle with the Missing Value Node.
BTW @KNIME-Team: the COL_MEAN(x) function in the Math Formula Node can not handle NaN values. Would be great if there would be an option to skip these values.