How to bulkload Agents

First open the agent bulkloader. This may take a while as it refreshes some cached data.

screen shot 2018-04-26 at 8 54 33 am

Click Agent Loader Home to access more tools

screen shot 2018-04-26 at 8 54 48 am

If you have a list of names

screen shot 2018-04-26 at 8 55 23 am

you can split them with the Agent Name Splitter.

screen shot 2018-04-26 at 8 54 55 am

Format is given at the top:

screen shot 2018-04-26 at 8 55 46 am

Upload, follow the link

screen shot 2018-04-26 at 8 55 46 am

This will run the agent checker function, and it will ALWAYS fail - agent type is not provided to this application in the interest of simplicity. Note any “fatal errors” - they will continue to cause problems.

screen shot 2018-04-26 at 8 56 50 am

The Agent Name Splitter injects a “formatted name.” You may wish to replace preferred_name with this standardization; you will need to keep data synchronized. In this case the preferred names were already formatted according to the Guidelines, there is nothing of value in “formatted name,” and it can simply be deleted.

screen shot 2018-04-26 at 8 58 43 am

The Agent Loader Home page provides a list of required and optional columns. We need to add agent_type - they’re probably all “person” if we successfully used the agent name splitter.

screen shot 2018-04-26 at 9 12 41 am

This is a good time to delete full duplicates as well; these agents exist and your data will already load against them.

screen shot 2018-04-26 at 9 02 16 am

back to the agent loader

screen shot 2018-04-26 at 9 00 09 am

Load the CSV you just created

screen shot 2018-04-26 at 9 00 42 am

follow the link

screen shot 2018-04-26 at 9 01 20 am

validate

screen shot 2018-04-26 at 9 01 36 am

back home

screen shot 2018-04-26 at 9 03 38 am

You will probably see an error message with an explanation. (Short version: Duplicate agents cause huge problems with data reliability; a little work now is MUCH easier than any alternative.)

screen shot 2018-04-26 at 9 04 05 am

An active member of the Arctos Working group (=has manage_codetables access) needs to review any errors and then may load these. The review process should include conformation that all information is being provided - e.g., that “J. S. Smith” is in these data because that’s absolutely all we know, NOT because tradition mandates entering “John Smithson Smith” as initials only.

Someone with admin access will see a scary red box with an override link.

screen shot 2018-04-26 at 9 04 48 am

Click it, if there are no remaining errors the data should load.

screen shot 2018-04-26 at 9 04 59 am

Related Links

How to Understand Bulkloader Results