This was causing NSUs to show a social care id. This now resolves this.

This reverts commit 004e831.

@SwiftySalmon spotted an issue where NSUs were showing a social care id
attached. This was due to us matching on NSUs first and then SC Client
variables at the end. Previously we matched SC client variables first
and then NSUs onto the episode file. This difference occurred when we
split up matching on client variables at the year specific extracts.
This is a quick PR to reorder the functions to match our previous logic.
I have tested the 1920 file to check this and the issue is no longer
there.

Improved the function `create_demog_test_flags` function to include a
parameter for specifying `chi` or `anon_chi`. This is due to this
function being used for testing the extracts (which calls chi) and also
being used for testing the episode and individual file (which calls
anon_chi). This now allows flexibility for interchanging between the two
forms and simplifies the code by removing duplicated code.

Closes #850

Checked for snake case and made some changes to the workbook names. I
have also updated the date stamp to be lower case as well.

Closes #508

This was not taking the correct variables, leading to NSUs being assigned psychiatry

closes #603

* Style code

* # read in sc demographics

different variables - removed extract date as not accurate, using chi over upi after discussion with social care data management. Added in date of death just for fun.

* social care demographics first draft

removed a lot of the submitted variables and instead using chi variables from chi seeding. Other changes:
- Fill in missing values,
- create flag for latest social care id (one from database is not accurate), this makes sure that each chi only has ONE sc id as the latest to stop it creating duplicates
- change postcode to choose chi over submitted

* Style code

* had a github error? Not sure what happened but commiting first draft of sc demographics

* Style code

* first draft sds.
No major changes - only how demographics is matched on and how latest social care id is selected

* Update documentation

* demographics - add sending location to group by

* Style code

* Update documentation

* Added ungroup()

* Remove comments

* Remove comments

* Style code

---------

Co-authored-by: SwiftySalmon <[email protected]>
Co-authored-by: marjom02 <[email protected]>
Co-authored-by: Jennit07 <[email protected]>
Co-authored-by: Jennit07 <[email protected]>
Co-authored-by: Zihao Li <[email protected]>

* speed up process_sc_all_alarms_telecare function with data.table package

* Update documentation

---------

Co-authored-by: lizihao-anu <[email protected]>
Co-authored-by: Megan McNicol <[email protected]>
Co-authored-by: Jennit07 <[email protected]>

…FALSE` (#911)

Add case_when statement for `high_cc` cohort

Due to a coding error, the `dplyr::pick` was not working correctly when
trying to take the average for assigning the label `unassigned`. This
was causing NSUs to have the label `psychiatry` attached as this was the
next name available. The fix is to specify all of the columns needed to
take an average for assigning the label `unassigned`. This has a big
impact on the figures.

The  current 'live' files:

```
  service_use_cohort       n
   <chr>                <int>
 1 Elective Inpatient  488510
 2 Geriatric           293863
 3 Limited Daycases    725538
 4 Maternity           206488
 5 Multiple Emergency 1140659
 6 Outpatients        1102837
 7 Prescribing        3982957
 **8 Psychiatry         1601193**
 9 Routine Daycase     178739
10 Single Emergency   1116014
**11 Unassigned          250917**
12 Unscheduled Care   1718283
13 NA                   55697

```

after the fix: 

```
   service_use_cohort       n
   <chr>                <int>
 1 Elective Inpatient  488521
 2 Geriatric           293869
 3 Limited Daycases    725583
 4 Maternity           206488
 5 Multiple Emergency 1140688
 6 Outpatients        1102878
 7 Prescribing        3982958
 **8 Psychiatry          111859**
 9 Routine Daycase     178742
10 Single Emergency   1116042
**11 Unassigned         1740237**
12 Unscheduled Care   1718269
13 NA                   55180
```

Ive asked the source team if they can run through a test to see the
impact on their figures so i am opening this as a draft for now incase
further changes are needed.

closes #907

* Add missing sc variables for no sc data

* Fix code for including `_inc_dna` variables

* Remove commented line

…ning. (#913)

Fix bug - pop file paths breaking indiv file

Opening a draft PR for now. I think we need to review the care home
methodology before deciding if we need the quarter calculation
functions. I will highlight areas in the code which i am not sure about
as i want to make sure we pick up the correct logic.

For now i have concluded that we can remove the `check_qtr_format`
function so have removed this from the code.

Closes Issue #886

A small issue was identified when running targets. Linked with changes to the function `fix_sc_end_dates()`

* fix sc_client_lookup sc_send_lca

* fix an issue of get_pop_path

* Style code

* fix the rest of get_pop_path from get_datazone_pop_path

* Update documentation

* fix sc_send_lca

* add missing year column

* explicitly specify the argument year to avoid corruption of targets

* Update documentation

* new data pipeline with targets
remove create_individual_files from targets and append it to run_targets script

* minor changes

* Style code

* undo sc_send_lca bit

* Update targets scripts

* Remove top level targets scripts

---------

Co-authored-by: lizihao-anu <[email protected]>
Co-authored-by: Megan McNicol <[email protected]>
Co-authored-by: Jennit07 <[email protected]>
Co-authored-by: Jennifer Thom <[email protected]>

A small issue was identified when running targets. Linked with changes
to the function `fix_sc_end_dates()`

Merge branch 'add_activity_after_death' of https://github.com/Public-Health-Scotland/source-linkage-files into add_activity_after_death
# Please enter a commit message to explain why this merge is necessary,
# especially if it merges an updated upstream into a topic branch.
#
# Lines starting with '#' will be ignored, and an empty message aborts
# the commit.

Co-authored-by: Jennit07 <[email protected]>

Co-authored-by: Jennit07 <[email protected]>