Skip to contents

Do some checking, sanitizing and write to backend database db_datawrite() does some checking, sanitizing and writes new data to backend database

Usage

db_datawrite(
  metadf,
  datadf,
  construct,
  grp = FALSE,
  engine = "postgresql",
  sanitize = TRUE,
  forceunique = T,
  replace = FALSE
)

Arguments

metadf

dataframe list with the following expected structure : [1] -> df with orig_name data_name data_desc [2] -> data.frame with at least data_class_id data_freq_id dataunit_num - TBI dataunit_den - TBI datasource_id data_url <- construct ? call? @param datadf dataframe with the following expected structure: local <-> local_id or name/code pairable with local table periodo <- date or string/number in yyyy,yyyy-mm, yyyy-mm-dd valor <- numeric @param construct The call that extracted/obtained the data @param grp Should a datagroup be defined - for tables with subgroups (as sidra) @param engine Defaults to postgresql, sqlite TBI.