Questions on WIP code in NewV1

[katy-sadowski]

I've starting working line by line through the code to inform continued work on the cohort line item class + associated table shell / SQL functionality. I have questions about some things which seem to be works in progress, and want to make sure I understand the plan/intention for them in order to stay consistent. I'll keep adding stuff to this issue as I work, and will probably end up figuring some of this out myself.

ClinicalCharacteristics/R/utilities.R

Line 427 in 5c5a61c

'distributionType' = purrr::map_chr(demoCatLi, ~.x$identifyStatType())

identifyStatType was defined in an R6 Class that's now commented out. Do we still want/need this method, or do we want to change the overall approach here?

ClinicalCharacteristics/R/R6Class.R

Line 380 in 5c5a61c

~.x$getSql()

getSql was defined in the commented-out demographics class. Do we still want/need this method, or do we want to change the overall approach here?

ClinicalCharacteristics/R/R6Class.R

Line 158 in 5c5a61c

private$.buildConceptLineItemQuery(),

It seems this should probably actually be buildConceptSetOccurrenceQuery?

[mdlavallee92]

Many of these things are being phased out since changing the approach of making a tsMeta table to orient the construction of sql. The distributionType for example will be removed when the process is complete. The getSql command is too generic and will likely be replaced with something more explicit to the class. Your last item yes that should be changed. I have not finished fixing all of these aspects yet.

The idea now is to build a meta table that orchestrates the table shell design. See...tableShell$getTableShellMeta(). The output of this function gives all the information needed to construct the required queries. In terms of cohort construction you need to track the cohort ids needed. You are right this is WIP and haven't had a good focus period to get over this hurdle.

[katy-sadowski]

Thanks for the input, that's very helpful and all makes sense! And no worries at all - I was just checking on these so I made sure to follow the right patterns for cohort. I'll comment back here for guidance if/when I get time to continue chipping away at this :)