sleep

[caitlink12]

sleep variables SLP_02, SLP_02_A, SLP_03, SLP_03_A, SLP_04, SLP_04_A, SLPG01_A, SLPG01_B, SLPG01_cont were harmonized into the variable and variable_details sheets for master survey cycles 2001, 2011-2012, 2013-2014, and 2015-2016.
master survey cycles 2011-2012, 2013-2014, and 2015-2016 were added to existing variables SLP_02, SLP_3, SLP_04.
master survey cycle 2001 was added to existing variables SLP_02_A, SLP_03_A, SLP_04_A.
master survey cycles 2001 and 2015-2016 were added to existing variable SLPG01_A
SLPG01_B was created for master survey cycles 2011-2012 and 2013-2014
master survey cycles 2001, 2011-2012, 2013-2014, and 2015-2016 were added to existing variable SLPG01_cont

[DougManuel]

The PR correctly adds Master file cycles. Passes all tests, including integration testing.

Sleeping variables are optional content, which means the availability varies from cycle to cycle and from province to province.

This PR changes:

• Adds Master file cycles (2001, 2011-2012, 2013-2014, 2015-2016) to sleep variables
• Creates SLPG01_B for Master cycles 2011-2014
• Correctly maps source variables across eras

Key findings

Optional content warning: Sleep module was Optional Content from 2007-2018. Provincial availability varies by cycle - not all provinces have sleep data in all cycles. Added description notes to all 11 sleep variables to alert users.

Scale incompatibility:

• 2001 uses 3-point scale (Most/Sometimes/Never) with inverted direction (lower = worse)
• 2007+ uses 5-point scale (None to All) where higher = worse
• Correctly handled via _A suffix variables for 2001

Cycles without sleep data (Canada-wide PUMF):

• 2003, 2005, 2009-2010 - correctly excluded from worksheets

Verification

• ✓ Variable naming evolution verified across all eras
• ✓ Missing cycles correctly excluded
• ✓ Scale incompatibility handled with separate _A variables
• ✓ Added optional content notes to variables.csv

Documentation

Full review documentation: https://dmanuel.quarto.pub/cep-005-sleep-variables

Notes

Added notes about data availability

[DougManuel]

Code review

Reviewed 10 sleep variables (SLP_02, SLP_02_A, SLP_03, SLP_03_A, SLP_04, SLP_04_A, SLPG01, SLPG01_A, SLPG01_B, SLPG01_cont) for PUMF and Master across 2001 through 2017-2018.

Note: This PR was previously approved (2026-01-27) but 2 commits were pushed after approval, triggering a re-review.

L6 integration test: cross-cycle prevalence

rec_with_table() ran successfully for all PUMF cycles. Sleep is optional content so valid percentages vary by cycle (provincial availability).

Variable	cchs2001_p	cchs2007_2008_p	cchs2011_2012_p	cchs2013_2014_p	cchs2015_2016_p	cchs2017_2018_p
SLP_02	-	3.0%	33.5%	3.0%	40.5%	33.0%
SLP_02_A	98.5%	-	-	-	-	-
SLP_03	-	3.0%	33.5%	3.0%	40.5%	33.0%
SLP_03_A	98.5%	-	-	-	-	-
SLP_04	-	3.0%	33.5%	3.0%	40.5%	33.0%
SLP_04_A	98.5%	-	-	-	-	-
SLPG01	-	-	33.5%	3.0%	40.5%	33.0%
SLPG01_A	98.5%	3.0%	-	-	-	-
SLPG01_cont	98.5%	3.0%	31.5%	7.0%	39.5%	46.0%

Master (_m) mappings validated by worksheet checks only -- no runtime data available for L6 testing. SLPG01_B is Master-only (cchs2011_2012_m, cchs2013_2014_m).

Issues found

3 PR-introduced issues. 6 pre-existing issues noted but not blocking.

PR-introduced:

1. SLP_04 else row missing Master databases (L3, confidence 90)
   The else catch-all row for SLP_04 has databaseStart = PUMF + cchs2012_s only, but all other rows (recStart 1-5, 6, [7,9]) include cchs2011_2012_m, cchs2013_2014_m, cchs2015_2016_m. This means unexpected values on Master databases won't be caught by the else->missing mapping.
   Fix: add cchs2011_2012_m, cchs2013_2014_m, cchs2015_2016_m to the else row's databaseStart.

2. variables.csv missing _m databases for 4 variables (L3, confidence 95)
   Master databases were added in variable_details.csv but not reflected in variables.csv for:

   • SLP_04_A: missing cchs2001_m
   • SLP_04: missing cchs2011_2012_m, cchs2013_2014_m, cchs2015_2016_m
   • SLPG01_A: missing cchs2001_m, cchs2015_2016_m
   • SLPG01_cont: missing cchs2001_m, cchs2011_2012_m, cchs2013_2014_m, cchs2015_2016_m
     This will prevent rec_with_table() from finding these variables when processing Master databases.

3. SLP_02/SLP_03: cchs2012_s vs cchs2012_m inconsistency (L3, confidence 85)
   variables.csv lists cchs2012_m but variable_details.csv still has cchs2012_s for SLP_02 and SLP_03. The two files should agree. If the intent is Master, both should use cchs2012_m.

Pre-existing (not blocking, for information):

• SPLG01_C retains deprecated _i databases (cchs2001_i, cchs2015_2016_i)
• SLPG01_B dummyVariable for [97,99] row uses NA::a instead of expected NA::b
• SLPG01_cont has typo "convereted" in catStartLabel
• SLPG01_C listed in variables.csv but has 0 rows in variable_details.csv (orphan entry)
• cchs2012_s (deprecated _s suffix) present across SLP_02, SLP_03, SLP_04
• variable_details.csv has 19 trailing empty columns from Excel editing (41 columns vs 22 expected)

GHA status

R-CMD-check failed with 2 ERRORs and 6 WARNINGs. These appear to be pre-existing package-level issues (undefined exports, parse errors) not caused by the sleep worksheet changes. The NAMESPACE commit (7cb75eb) partially addressed this but issues remain.

Checked

Era boundary defaults, databaseStart consistency (variables.csv vs variable_details.csv), PUMF vs Master naming, deprecated _s/_i suffixes, dummyVariable naming, label consistency, and PUMF integration testing.

CEP: ceps/cep-005-sleep/

[DougManuel]

All 3 PR-introduced issues from the review have been fixed in daaec8b:

1. SLP_04 else row: added cchs2011_2012_m, cchs2013_2014_m, cchs2015_2016_m
2. variables.csv: added missing _m databases for SLP_04, SLP_04_A, SLPG01_A, SLPG01_cont
3. cchs2012_s → cchs2012_m: replaced deprecated _s suffix in SLP_02, SLP_03, SLP_04 (both files)

Ready to merge. @rafidoole for final checks.

[rafdoodle]

Final SLP Variable Changes

Overview

Action	Variable	Notes
Renamed	SLP_02_A → SLP_02_cat3	Name now reflects recoding type
Renamed	SLP_03_A → SLP_03_cat3	Name now reflects recoding type
Renamed	SLP_04_A → SLP_04_cat3	Name now reflects recoding type
Renamed	SLPG01 → SLPG01_pre2015	Scoped to pre-2015 PUMF cycles
Renamed	SLPG01_A → SLPG01_cat12	Name now reflects 12-category scale
Added	SLPG01_2015plus	New variable for 2015+ PUMF cycles
Deleted	SLPG01_B	Superseded by SLPG01_pre2015
Deleted	SLPG01_C	Orphan variable (no rows in variable_details.csv); removed
Modified	SLP_02, SLP_03, SLP_04	databaseStart/variableStart expanded
Modified	SLPG01_cont	databaseStart/variableStart expanded; label fixes

---

variables.csv changes

Variable	Field	Before	After
SLP_02	databaseStart	PUMF cycles only	Added master file cycles
SLP_03	databaseStart	PUMF cycles only	Added master file cycles
SLP_04	databaseStart	PUMF cycles only	Added master file cycles
SLPG01_pre2015 (was SLPG01)	databaseStart	All cycles	Pre-2015 PUMF cycles only
SLPG01_cat12 (was SLPG01_A)	databaseStart	2015+ cycles	All cycles including cchs2001_p/m via GENA_03
SLPG01_cont	databaseStart	Subset of cycles	Expanded to full set of supported cycles
SLPG01_2015plus	—	(did not exist)	New; covers 2015+ PUMF cycles, 11-category scale
SLPG01_B	—	Present	Deleted
SLPG01_C	—	Present	Deleted (orphan — no corresponding rows in variable_details.csv)

---

variable_details.csv changes

Renames & dummyVariable fixes

Variable	Rows	dummyVariable pattern — Before	dummyVariable pattern — After
SLP_02_cat3 (was SLP_02_A)	6	SLP_02_A_cat3_*	SLP_02_cat3_*
SLP_03_cat3 (was SLP_03_A)	6	SLP_03_A_cat3_*	SLP_03_cat3_*
SLP_04_cat3 (was SLP_04_A)	6	SLP_04_A_cat3_*	SLP_04_cat3_*
SLPG01_pre2015 (was SLPG01)	14	SLPG01_cat11_*	SLPG01_pre2015_cat11_*
SLPG01_cat12 (was SLPG01_A)	15	SLPG01_A_cat12_*	SLPG01_cat12_cat12_*

New rows added

Variable	Rows	Description
SLPG01_2015plus	14	11-category structure mirroring SLPG01_pre2015; covers cchs2015_2016_p and cchs2017_2018_p via [SLPG005]

Rows deleted

Variable	Rows	Reason
SLPG01_B	14	Superseded by SLPG01_pre2015
SLPG01_C	0	Orphan variable; existed only in variables.csv with no recoding rows
SLPG01_cont block 3	14	Duplicate master-file recoding block; removed

databaseStart / variableStart updates

Variable	Change
SLP_02, SLP_03, SLP_04	Updated to match expanded variables.csv cycles
SLPG01_pre2015	Set to pre-2015 PUMF cycles
SLPG01_cat12	Set to all supported cycles including cchs2001_p/m
SLPG01_cont block 1	Updated to match SLPG01_cat12 cycles

Label corrections

Variable	Field	Before	After
SLPG01_pre2015, recStart = 1	catLabel, catLabelLong, catStartLabel	<3 hours	<2 hours
SLPG01_cont block 1, all 12 category rows	catLabel	Hours	Proper hour ranges (e.g., <2 hours, 2-<3 hours, …, >= 12 hours)
SLPG01_cont block 2, recStart = 1	catLabel	<3 hours	<2 hours
SLP_03 NA rows	dummyVariable	SLP_03_cat3_NA	SLP_03_cat5_NA

[rafdoodle]

Changed manually merged to v3 via commit 699dbff. Will close this PR now.