Alteryx Designer Desktop Discussions

Find answers, ask questions, and share expertise about Alteryx Designer Desktop and Intelligence Suite.
SOLVED

Association analysis Error

sraja
5 - Atom

Hello All,

I am using association analysis to find which features are more significant than others to contribute to the target variable. My target variable is a categorical field- 1 and 0. 1= Yes and 0=No

 

When I use association analysis, and choose this field as the Target field and add '1' in the target level of interest and select rest of the fields, I receive an error: Error: Association Analysis (36): Tool #9: The provided target level is not one of the possible field values.

 

The value of 1 does exist for the target variable. Also, I marked this variable as a String in the 'Select' tool.

 

I am not sure how to proceed. Any insight provided would be helpful.

Thanks.

2 REPLIES 2
PaulN
Alteryx Alumni (Retired)

Hi @sraja,

 

Just to confirm, do you specify '1' or 1 as the target level of interest? If target field is not numeric and only contains 0 or 1 then these will be the only valid values (no quotes or double-quotes needed).

 

Example:

 

association_analysis_1.png

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

With a valid target level (either 0 or 1):

 

association_analysis_2.png

 

association_analysis_3.png

 

 

With an invalid target level ('1' is not matching any VS values):

 

association_analysis_4.png

 


association_analysis_5.png

 

I hope this helps.

 

Thanks,

 

Paul Noirel

Sr Customer Support Engineer

sraja
5 - Atom

Thanks Paul for your reply. It was very helpful. I was using '1' and I was able to get my association analysis to work with your help.

 

However, I also noticed that due to lot of missing values, the association analysis was giving me errors. I fixed it using the Imputation tool.

 

One more question- if I have categorical feature, association analysis using chi-square would not work. What algorithm should I use for this, as I want to find the most significant feature, which may include this categorical field.

 

Thank you so much for your help.

Labels