"Data cleansing - float and text"

Katzmann

Hello everyone,

i have a data set written in an Excel table. In one column there are floats and text, e.g.

0,00
0,54
May-94
0,23

My first goal is to filter out the text and give a list of these errorneous datasets.
Second I would like to replace them, e.g. May-94 by 5,94.

First I try to read them via "Read Excel" giving them the text attribute. But then I am stuck. Turning them into nominals ("Text to Nominal") and numerics (Nominal to Numeric") did not help, since all the numbers are replace by 0,1,2,3....

Is there a way to do this in rapidminer?

Thanks,

Katzmann