Check Patterns in Data (using multiple fields)

[newsqluser02]

I am new to SQL and trying to find a more effective way to accomplish what I'm doing.

So I am trying to find patterns for the errors in my data. I have several fields that can determine if these errors exist.

Here are the fields that I have in my database

 company_code, deal_id, sex_code, unisex, issue_age, level_term_period, underwriting_class, rider_code, table_rating, policy_year, band, face_amount, premium_rate, calculated_premium_rate, premium_rate_variance

I am wanting group the data by company_code first and then see if one said company_code has more premium_rate_variance that has values "Yes" than values "No".

If this is not the case, break the grouping down further into company_code and deal_id. Then check which premium_rate_variance has more "Yes" than "No"

And continue breaking it down until it has sorted it out completely.

I know this is very basic, because to truly see patterns, I'd have to start with run each of the fields seperate, then run two fields, then three, etc.

I have written some code, though I know it is wrong but it shows where I am going:

IF 
 (select count(company_code) from treaty_validation_data where premium_rate_variance = 'Y' group by company_code) 
 >
 (select count(company_code) from treaty_validation_data where premium_rate_variance = 'N' group by company_code)
   select * from treaty_validation_data
   where (select count(company_code) from treaty_validation_data where premium_rate_variance = 'Y' group by company_code) 
          >
         (select count(company_code) from treaty_validation_data where premium_rate_variance = 'N' group by company_code)
   group by company_code
ELSE
 IF 
  (select count(company_code) from treaty_validation_data where premium_rate_variance = 'Y' group by company_code, deal_id) 
  >
  (select count(company_code) from treaty_validation_data where premium_rate_variance = 'N' group by company_code, deal_id)
    select * from treaty_validation_data 
    where (select count(company_code) from treaty_validation_data where premium_rate_variance = 'Y' group by company_code, deal_id) 
          >
          (select count(company_code) from treaty_validation_data where premium_rate_variance = 'N' group by company_code, deal_id)
    group by company_code, deal_id
ELSE
 IF
  (select count(company_code) from treaty_validation_data where premium_rate_variance = 'Y' group by company_code, deal_id, sex_code) 
  >
  (select count(company_code) from treaty_validation_data where premium_rate_variance = 'N' group by company_code, deal_id, sex_code)
    select * from treaty_validation_data
    where (select count(company_code) from treaty_validation_data where premium_rate_variance = 'Y' group by company_code, deal_id, sex_code) 
          >
          (select count(company_code) from treaty_validation_data where premium_rate_variance = 'N' group by company_code, deal_id, sex_code)
    group by company_code, deal_id, sex_code