DOC: Add columns to decennial census and ACS #229
Conversation
NathanielBlairStahn
added
the
documentation
NathanielBlairStahn
requested review from
a team,
zmbc and
pletale
as code owners
docs/source/datasets/index.rst
Outdated
| * - Relationship to reference person | ||
| - :code:`relationship_to_reference_person` | ||
| - :code:`relationship_to_reference_person` | ||
| - Possible values for this indicator include: |
There was a problem hiding this comment.
I noticed that the "relationship to reference person" column for the decennial census is missing some possible values, namely: "Opposite-sex spouse"; "Opposite-sex unmarried partner"; "Same-sex spouse"; "Same-sex unmarried partner"
I did not update these values in this pull request in case we want to make this correction as a hotfix instead. So if this PR gets merged to develop, the documentation for census and ACS will not mach until we implement the hotfix and merge it to develop. Alternatively, I could just make the change here, and we could not worry about making the correction in a hotfix. @aflaxman any preference?
There was a problem hiding this comment.
Keep it simple, just make the changes here.
| institutional". The types of noninstitutional group quarters are | ||
| "College", "Military", and "Other noninstitutional". | ||
| * - Relationship to reference person | ||
| - :code:`relationship_to_reference_person` |
There was a problem hiding this comment.
As discussed on slack, our data column in the decennial census is called relation_to_reference_person instead. Should I change this to match, or should we change the data to match what's in the docs?
There was a problem hiding this comment.
I'm leaning towards relationship_to_reference_person at this moment.
NathanielBlairStahn
changed the title
docs/source/datasets/index.rst
Outdated
| "Reference person"; "Opposite-sex spouse"; "Opposite-sex unmarried | ||
| partner"; "Same-sex spouse"; "Same-sex unmarried partner"; "Biological | ||
| child"; "Adopted child"; "Stepchild"; "Sibling"; "Parent"; "Grandchild"; | ||
| "Parent-in-law"; "Child-in-law"; "Other relative"; "Roommate"; "Foster | ||
| child"; "Other nonrelative"; "Institutionalized group quarters | ||
| population"; and "Noninstitutionalized group quarters population". |
There was a problem hiding this comment.
I used the full category names such as "Opposite-sex unmarried partner" and "Noninstitutionalized group quarters population" instead of our abbreviations such as "Opp-sex partner" and "Noninstitutionalized GQ pop". I think that's better -- do other people have a preference?
There was a problem hiding this comment.
agree, since the strings are long anyway we might as well have them unabbreviated.
There was a problem hiding this comment.
If we're changing the "GQ pop" ones anyway, did we decide whether we like the word "population"? I feel like "group quarters person" is more natural for this column but don't feel strongly.
docs/source/datasets/index.rst
Outdated
| "Reference person"; "Opposite-sex spouse"; "Opposite-sex unmarried | ||
| partner"; "Same-sex spouse"; "Same-sex unmarried partner"; "Biological | ||
| child"; "Adopted child"; "Stepchild"; "Sibling"; "Parent"; "Grandchild"; | ||
| "Parent-in-law"; "Child-in-law"; "Other relative"; "Roommate"; "Foster | ||
| child"; "Other nonrelative"; "Institutionalized group quarters | ||
| population"; and "Noninstitutionalized group quarters population". |
There was a problem hiding this comment.
If we're changing the "GQ pop" ones anyway, did we decide whether we like the word "population"? I feel like "group quarters person" is more natural for this column but don't feel strongly.
I got the names from pp. 39-40 in this reference, with the only change being to remove the gendered language: If we wanted to make other changes I guess we could, but I figured the simplest thing was to use what's there. |
Yes, that makes sense. More realistic to how the real data is, too. |
| @@ -6,10 +6,10 @@ Datasets | |||
|
|
|||
| Here we cover the realistic simulated datasets, which are analogous to "real world" administrative records such as tax documents | |||
| and routinely generated files of social security numbers, that users can generate using Pseudopeople for developing and testing Entity | |||
There was a problem hiding this comment.
| and routinely generated files of social security numbers, that users can generate using Pseudopeople for developing and testing Entity | |
| and routinely generated files of social security numbers, that users can generate using pseudopeople for developing and testing Entity |
I just realized the capitalization for 'pseudopeople' on this page is inconsistent with the rest of documentation - it should all be lower-case right?
There was a problem hiding this comment.
Yes, thanks, pseudopeople should be lower case. I think there are several minor edits like this that we should do to improve the documentation, but I haven't been prioritizing them since we're still working on getting the main content in line with what the engineers are doing. For the sake of limiting the scope of each PR, I think I'm not going to make this change in this PR, but I'll create a JIRA ticket to take a pass at editing the Datasets page. Note that there are a few more edits to make on this page, e.g., at least one more instance of capitalized "Pseudopeople", and "Entity Resolution" should also be lower case.
Co-authored-by: Zeb Burke-Conte <[email protected]>
Co-authored-by: Zeb Burke-Conte <[email protected]>
Mic 4255/relationship category name updates Updates categories of relationship to reference person among household members. - *Category*: Data - *JIRA issue*: [MIC-4255](https://jira.ihme.washington.edu/browse/MIC-4255) - *Research reference*: ihmeuw/pseudopeople#229 Changes and notes -updates relationshiop to reference person category names. Makes it so abbreviated names are no long abbreviated. Verification and Testing Successfully rebuilt data key in artifact. Ran simulation and make_results.
DOC: Add columns to decennial census and ACS
Description
This is intended to be the official documentation for the engineers to implement these changes, as well as being the user-facing documentation.
The diff includes a bunch of whitespace deletions from my editor, so you may want to hide those.
Testing
Built docs locally