R/RowsColsQ.R
rowrecs_to_blocks_q.Rd
Transform data facts from columns into additional rows using SQL and controlTable.
rowrecs_to_blocks_q(
wideTable,
controlTable,
my_db,
...,
columnsToCopy = NULL,
tempNameGenerator = mk_tmp_name_source("mvtrq"),
strict = FALSE,
controlTableKeys = colnames(controlTable)[[1]],
checkNames = TRUE,
checkKeys = FALSE,
showQuery = FALSE,
defaultValue = NULL,
temporary = FALSE,
resultName = NULL,
incoming_qualifiers = NULL,
outgoing_qualifiers = NULL,
temp_qualifiers = NULL,
executeQuery = TRUE
)
name of table containing data to be mapped (db/Spark data)
table specifying mapping (local data frame)
db handle
force later arguments to be by name.
character array of column names to copy
a tempNameGenerator from cdata::mk_tmp_name_source()
logical, if TRUE check control table name forms
character, which column names of the control table are considered to be keys.
logical, if TRUE check names
logical, if TRUE check wideTable keys
if TRUE print query
if not NULL literal to use for non-match values.
logical, if TRUE make result temporary.
character, name for result table.
optional named ordered vector of strings carrying additional db hierarchy terms, such as schema.
optional named ordered vector of strings carrying additional db hierarchy terms, such as schema.
optional named ordered vector of strings carrying additional db hierarchy terms, such as schema.
logical, if TRUE execute the query and return result.
long table built by mapping wideTable to one row per group (or query if executeQuery is FALSE)
This is using the theory of "fluid data"n (https://github.com/WinVector/cdata), which includes the principle that each data cell has coordinates independent of the storage details and storage detail dependent coordinates (usually row-id, column-id, and group-id) can be re-derived at will (the other principle is that there may not be "one true preferred data shape" and many re-shapings of data may be needed to match data to different algorithms and methods).
The controlTable defines the names of each data element in the two notations: the notation of the tall table (which is row oriented) and the notation of the wide table (which is column oriented). controlTable[ , 1] (the group label) cross colnames(controlTable) (the column labels) are names of data cells in the long form. controlTable[ , 2:ncol(controlTable)] (column labels) are names of data cells in the wide form. To get behavior similar to tidyr::gather/spread one builds the control table by running an appropriate query over the data.
Some discussion and examples can be found here: https://winvector.github.io/FluidData/FluidData.html and here https://github.com/WinVector/cdata.
if (requireNamespace("DBI", quietly = TRUE) &&
requireNamespace("RSQLite", quietly = TRUE)) {
my_db <- DBI::dbConnect(RSQLite::SQLite(), ":memory:")
# un-pivot example
d <- data.frame(AUC = 0.6, R2 = 0.2)
rquery::rq_copy_to(my_db,
'd',
d,
overwrite = TRUE,
temporary = TRUE)
cT <- build_unpivot_control(nameForNewKeyColumn= 'meas',
nameForNewValueColumn= 'val',
columnsToTakeFrom= c('AUC', 'R2'))
tab <- rowrecs_to_blocks_q('d', cT, my_db = my_db)
qlook(my_db, tab)
DBI::dbDisconnect(my_db)
}
#> table `mvtrq_58589036583724552768_0000000001` SQLiteConnection
#> nrow: 2
#> 'data.frame': 2 obs. of 2 variables:
#> $ meas: chr "AUC" "R2"
#> $ val : num 0.6 0.2