7,649
edits
Changes
Jump to navigation
Jump to search
→Fund and Branch
===Fund and Branch===
We have to key off of fundname for the fund table. Fundbase has 31931 on load. Throwing out undisclosed funds leaves 31540 with 31479 distinct names (requiring disclosed firmnames too reduces this to 31404). The solution is to keep the duplicate with the largest numportcos, discarding Net Ventures, Cambridge Innovations - Unspecified Fund', and Spring Capital, as these are the only three with a tie and both are of no consequence (numportco=1).
Likewise, we have to key off of firmname for all firm tables. It isn't that we don't have other identifying information in firmbase (all records have founding date) but we only have firmname as a foreign key in the fund table and branch office table.
