minor clean up of working file

R/abc.R

@@ -10,14 +10,14 @@
 #' @param b incremental segment threshold expressed as a percentage eg .26
 #' @param c final segment threshold expressed as a percentage eg .04
 #' @param func either "sum" or "n"; sum to sum your dim or n for row count
-#' @param ... columns  to group by
+#' @param ... columns to group by
 #'
 #' @return a tibble or DBI object
 #' @export
 #'
 #' @examples
 #'
-#' customer_abc <- abc(fpaR::Sales,customer_key,dim=quantity,a=.7,b=.2,c=.1)
+#' customer_abc <- abc(fpaR::sales,customer_key,dim=quantity,a=.7,b=.2,c=.1)
 abc <- function(.data,...,dim,a=.7,b=.26,c=.04,func=c("sum")){
 
   ## function check

R/cohort_date.R

@@ -7,15 +7,15 @@
 #'
 #' @param .data Either a database or tibble object
 #' @param id_var The variable that you want to track over time
-#' @param time_unit The time unit as character eg('day','week','month',etc) to round the transaction date passed to lubridate::floor_date()
+#' @param time_unit The time unit as string eg('day','week','month',etc) to round the transaction date passed to lubridate::floor_date()
 #' @param date_var  The date column representing the transaction date
 #' @param period_label Logical value if you want the represent the dates as periods or Dates
 #'
 #' @return tibble or DBI object
 #' @export
 #'
 #' @examples
-#' fpaR::contoso_fact_sales |>  make_cohort_tbl(id_var=ProductKey,date_var=DateKey,time_unit = 'week',period_label =TRUE)
+#' fpaR::sales |>  make_cohort_tbl(id_var=product_key,date_var=order_date,time_unit = 'week',period_label =TRUE)
 #'
 #'
 make_cohort_tbl <- function(.data,id_var,date_var,time_unit="month",period_label=FALSE){

R/show_in_excel.R

@@ -7,7 +7,7 @@
 #' @export
 #'
 #' @examples
-#' ggplot2::diamonds |>  show_in_excel()
+#' mtcars |>  show_in_excel()
 show_in_excel <- function(.data){
 
 # supporting arguments

R/time_intelligence.R

@@ -72,6 +72,69 @@ make_aggregation_tbl <- function(.data,...,date_var,value_var, time_unit) {
 
 }
 
+
+
+
+make_aggregation_sql <- function(.data,...,date_var,value_var, time_unit){
+
+ cols <-  enquos(...)
+ group_cols <- deparse(cols)
+
+
+  summary_sql <- glue_sql("
+  WITH summary_tbl AS (
+    SELECT
+      DATE_TRUNC({time_unit}, {`date_var`}) AS date,
+      {`value_var`},
+      {`glue_sql_collapse(group_vars, sep = ', ')`}
+    FROM {`DBI::SQL(.data)`} -- Replace .data with your actual table
+    GROUP BY date, {`glue_sql_collapse(group_vars, sep = ', ')`}
+  ),
+
+  calendar_tbl AS (
+    SELECT
+      generate_series(
+        DATE_TRUNC({time_unit}, MIN(date)),
+        DATE_TRUNC({time_unit}, MAX(date)),
+        INTERVAL '1 {time_unit}'
+      )::DATE AS date
+    FROM summary_tbl
+  ),
+
+  crossing_tbl AS (
+    SELECT
+      calendar_tbl.date,
+      {`glue_sql_collapse(group_vars, sep = ', ')`}
+    FROM
+      calendar_tbl
+      CROSS JOIN (
+        SELECT DISTINCT {`glue_sql_collapse(group_vars, sep = ', ')`}
+        FROM summary_tbl
+      ) AS distinct_groups
+  ),
+
+  full_tbl AS (
+    SELECT
+      crossing_tbl.date,
+      {`glue_sql_collapse(group_vars, sep = ', ')'},
+      COALESCE(summary_tbl.{`value_var`}, 0) AS {`value_var`}
+    FROM
+      crossing_tbl
+      LEFT JOIN summary_tbl
+      ON crossing_tbl.date = summary_tbl.date
+      AND {`paste(glue_sql('{distinct_groups}', .con = DBI::SQL('')), collapse = ' AND ')`}
+  )
+  SELECT * FROM full_tbl
+", .con = DBI::SQL("your_database_connection"))
+
+}
+
+
+
+
+
+
+
 #
 # calculate <- function(.data,...,date_var,value_var,time_unit){
 #