Carrier and Policy Mapping being incomplete

[GloveBoxSean]

The Carrier & Policy mapping tables are incomplete or otherwise messy, so the GREL .cross() function doesn't return values for each cell in the 1000-generic_insurance_company_policy_report file, even with a few straightforward standardizations like trim() and toLowercase(). I see about 500 missing values by filling glovebox_policy_id crossed on Line_of_Business, and 23 missing by filling glovebox_carrier_id crossed on Writing_Company.

It looks like the source values for these missing cells aren't listed in the mapping tables, e.g. "Auto (Personal)" or "RT SPECIALTY INS". Based on what you showed me during our demo call, in live work we'd add these values to the mapping tables keyed to their appropriate target values. How do you want me to handle these cases now?

I should be able to puzzle out how these missing values map based on context and could modify the mapping tables to be complete for this case, raising issues for any value where the correct mapping isn't clear. Do you have a process for documenting these mapping changes via GitHub you want me to follow?

[GloveBoxSean]

For the Policy type mapping, make sure you are reviewing all available columns. There is a better column to use than the 'Line of Business'. Take a look at this issue for more detail: [https://github.com//issues/2]

For the carrier mapping: What we have done is to do a transformation on the 'carrier_name' column to convert those records to a new column that is all lower case. Then do the mapping based on this column as the unique identifier on the 'ams_carrier_name_mapping' file. This should line you up.

It is totally normal for some to return blanks, since we do not map all carriers right now

Question about the 'writing_carrier': this is a nuance within the Insurance space. There are often times different companies that handle the writing of the insurance policy. There is a policy holder facing one and an internal facing one. This happens when a large carrier buys another one. We do not map based on this column, we only use it in our full data cleaning to gain more inside into what carrier this could be. So this is not needed in this challenge