Decumulate accumulated variables

Many models store accumulated values as the amount accumulated since the start of the forecast. decum() decumulates those values to specific time periods. Note that the time periods are overlapping, such that for hourly data a (for example) 6-hour accumulation will be calculated every hour for the previous 6 hours rather than in consecutive 6-hour windows.

Usage

decum(.data, decum_time, cols, time_col, group_col, df_name = NULL, ...)

Arguments

.data

A harp_df data fram or harp_list

decum_time

The time period to which to decumulate the data. If numeric, it is considered to be in hours, otherwise a string with a number followed by a unit. Units can be "s", for seconds; "m", for minutes; "h", for hours; or "d", for days.

cols

Columns for which to do the decumulation. Uses the same semantics as select.

time_col

The time column to use for calculating decumulation windows. Uses the same semantics as select.

group_col

The columns to group the output by. Uses the same semantics as select.

df_name

The name of the data frame - only really used to generate error messages.

...

Used for methods.

Value

An object of the same class as .data containing the decumulated values.