Workflow update

[seancsmith]

R session keeps crashing due to the size of the pico tag table in stroke-SOLES.

Have a look at how these tables are put together. Possibly take apart the pico_ont by type first e.g.

pico_ont_outcome <- dbreadtable(con, "pico_ontology)
filter(type == "outcome")

pico_tagged_outcome <- tbl(con, "pico_tag)
filter(regex_id %in% pico_ont_outcome$regex_id)

Rather than bringing in the full pico_tagged table!?