What it does | Motivation | Installation | Intended use case | make_test_dat() makes either a toy data.frame or database table in memory with known number of rows that satisfy the query we will show later | this is a database table | note that in-memory SQLite database stores dates as numbers | this is a local data.frame/tibble | convert Date to numeric to be consistent with claim_db | Interactive functions | here we use filter mode to remove records that failed the restriction | here we use flag mode to flag records that met the restriction instead of removing those | Class of result4 | execute query and download the result | Number of rows in source | Number of rows in the current result | make two look up tables | get year from dates for matching | note that keys must be present in all tables | Call-building functions | Next steps

Report sample sizes for your flowchart with report_n() | Compute age and age groups with compute_duration() | Decode codes with lookup() | Score comorbidity with compute_comorbidity() | Combine related records into episodes with collapse_episode() | Split long periods into regular intervals with cut_period() | Check date patterns directly with if_date() | Where these fit in a typical workflow

Why not brute force? | The easy half: within | The hard half: apart | Anchor at the extremes | A worked example | Both conditions at once | Teaching SQL to loop | Trust, but verify | Where this sits among known techniques