Due to recent updates, all users are required to create an Altair One account to login to the RapidMiner community. Click the Register button to create your account using the same email that you have previously used to login to the RapidMiner community. This will ensure that any previously created content will be synced to your Altair One account. Once you login, you will be asked to provide a username that identifies you to other Community users. Email us at Community with questions.
Replace (dictionary)
Contributor I
I have a data set that has 30+ attributes. Each data row has numerical codes in each column that correlate to a classification. For example; Gender is an attribute with codes 1-3 meaning Male, Female, and not provided, respectively. There are similar code structures for ethnicity, race, etc. I have set up a dictionary for each one of these attributes so that my model can reference to the specific dictionary and convert the codes to meaningful data. I have 2 questions:
1): Codes mean something different for each attribute conversion I'm performing so I set up separate dictionaries for each. For instance 1 means male for gender but it also means white for race and single for marital status. Is there a way to use the loop operator to have RM run all 30+ conversions using the different dictionaries or do I need to have 30 separate "Replace (Dictionary)" operators in my process?
2): In some dictionaries there are layered codes, for instance in my use case
1 = Latino/Hispanic
4 = N/A
14 = Other Hispanic or Latino
Instead of returning "Other Hispanic or Latino" for codes that equal 14, the operator is returning "Latino/HispanicN/A". I've seen that the regular expression option could prevent this however since I have the operator set up to run on a subset (the various ethnicity related attributes) and I do not want it applied to the whole population, I'm not sure that'd work. How can I go about fixing this?
1): Codes mean something different for each attribute conversion I'm performing so I set up separate dictionaries for each. For instance 1 means male for gender but it also means white for race and single for marital status. Is there a way to use the loop operator to have RM run all 30+ conversions using the different dictionaries or do I need to have 30 separate "Replace (Dictionary)" operators in my process?
2): In some dictionaries there are layered codes, for instance in my use case
1 = Latino/Hispanic
4 = N/A
14 = Other Hispanic or Latino
Instead of returning "Other Hispanic or Latino" for codes that equal 14, the operator is returning "Latino/HispanicN/A". I've seen that the regular expression option could prevent this however since I have the operator set up to run on a subset (the various ethnicity related attributes) and I do not want it applied to the whole population, I'm not sure that'd work. How can I go about fixing this?
Tagged:
0
Best Answer
-
kayman
Member Posts: 662 
Unicorn
That's because it is first replacing 1 and then 4 rather than using 14
Changing the order (so putting 14 before 1 and 4) may fix it already. When using regex you can force to have a full match, but the standard one replaces basically anything that matches, so 14 is considered as 1 and 4.7
