How to index and find common elements in the list ?

Harshav

I found that if a cell in an attribute has list of elements .We can index only one element using [] .But what if we want to index all the elements like [0:] in python. 
Finally, I want to find out common elements of two attributes in which all rows contain list data .

ceaperez

Hi @Harshav, 

please check the Select operator:
Select - RapidMiner Documentation

Also you can work with Python using  the Execute Python operator. 
please find attached a example process for that.

Best

Harshav

There is no code policy in my project and I see I cant loop list like data of each cell something like this  [1,2,3,45,6] 

BalazsBarany

Hi @Harshav,

there is a Split operator that you can use to split up the contents of an attribute on the specified separator character.

Usually you would then use De-Pivot to convert the newly created columns (attribute_1, attribute_2 etc.) to rows and process them with Aggregate (to get the most frequent values etc.).

Regards,
Balázs

MartinLiebig

Hi Harshav,

I am a bit confused here. Also pd.Dataframe does not allow a list in a cell. So the python equavalent does not really hold here. As mentioned, the split operator is of course the usual way to go here.

Best,

Martin

Harshav

Split operator gives you the result in that form ,but here I have compare two split values that are in this form["..."]

BalazsBarany

I don't understand your question.
Can you write examples of the data you have to process?

Are the brackets there before the split? 
[val1,val2,val3]
Then use Replace to remove the brackets with a regular expression.

Regards,
Balázs

Harshav

Okay...If we remove[] using reg exp , How do we compare val1,val2,val3 with another attribute which contains val2,val4,val5.?I want this val2 element to pop up.

BalazsBarany

If you de-pivot the data, your val1, val2, ... val5 will be simple values in rows. You can then use Join on a key field and the value to get the common elements.