Why is the new Rapid Miner 9.4 removing some content for my excel document after importing.

AmosGH

I was importing an excel file into the new RAPID MINER 9.4 repository but after loading it, it removed most of my content before the hashtag.
My old Rapid miner was not giving me this issue. Help

lionelderkrikor

Hi@AmosGH,

There is in deed a bug inRead Excel operator when there are return carriages (\r) and jump to the line (\n) in your initial excel file.
我已经描述了这个bugprevious thread.
After experimentations and "a little luck", the workaround here is to replace the "return carriages" and the "jump to the line" by .... nothing via the operatorReplacejust afterRead Excel.



The process in attached file.
PS :@sgenzer. I suggest to push this thread in "Bug Report" if the RM developpers are not already aware of this "bug".
Thanks you,

Regards,

Lionel

sgenzer

done. Thank you@lionelderkrikor

AmosGH

Thank you, sir, your reply was helpfulbut aside writing the expression in the replace operator is there a permanent solution looking forward.

lionelderkrikor

Hi@AmosGH,

A priori, actually there is only this workaround (via theReplaceoperator).
There is no permanent solution actually, that's why Scott submitted this to RapidMiner's development team : in order to find a permanent solution (fix the problem).

Regards,

Lionel

AmosGH

Thank you

alebo

Hi@AmosGH,
The content is still there, it's just visible when you hover over the field. What is the expected behavior? We don't change the content when reading a file. I would consider this to be a feature request.
Regards,
Andras

lionelderkrikor

Hi Andras,

I think there is still a problem with theRead Exceloperator associated withExecute Pythonoperator when there is text in an example set :
When a such example set enters in theExecute Pythonoperator, it seems that some examples are "splitted" in several examples.

In entry of the Execute Python, and so after theRead Exceloperator i have the following exampleset :



After the Execute Python operator , the example set becomes an example set with 122 examples : (The Execute Python does nothing)

I think this behaviour is linked to the "carriage return" and "line feed" because if we put aReplaceoperator with the setting (described in my previous post) just before the Execute Python operator, all is allright.

You can reproduce this behaviour by running the process in attached file (Enable/disable theReplaceoperator to see the difference).
The Excel file (data) is provided by@AmosGHin a previous post.

Regards,

Lionel

gmeier

Hi@lionelderkrikor,

you are right, the Execute Python operator has a problem with multi-line Strings. We are already working on fixing this.

However, this is not connected to the Read Excel operator. The excel file has entries with more than one line.

Best,

G

lionelderkrikor

Hi@gmeier,

OK, thanks you for your reply.

Regards,

Lionel