MissingValueReplenishment

laurab

Hi all,

Could anyone please tell me how the "average" function on the MissingValueReplensihment is calculated?

Is it the mean average?

Using the average function, does it average :

1. accross "ALL" the data i.e all rows and all columns
2. accross current row
3.accross current column

If there are several missing values, after the first missing value is replaced, does rapidminer include this value that has been replaced for calculating the next missing value. Or when a missing value is replaced does it not get used to calculate the remaining missing values. 

Cheers
Laura

land

Hi Laura,
the behavior of this operator should be very easyly recognized if you would apply it.
But here is what it does: A missing value in an attribute will be replaced by the selected function calculated over all non-missing values of this attribute. This is done for every attribute. This inserted values are not used for calculating the remaining values, but if you use the average, this shouldn't make a difference at all.

Greetings,
  Sebastian

laurab

Hi Sebastian,

Thanks for you reply it was really helpful.

Cheers Laura  8)