fill_na: fill missing values in the data of a pollyvote object

In pollyvote/pollyvoter: Aggregate Components, Calculate Forecasts and Determine Errors for Election Results.

Description

internal function that handles missing values in the data frame of a pollyvote object. Different error handling arguments can be handed over.

Usage

fill_na(dat, na_handle = c("last", "omit", "mean_within", "mean_across"),
  pv = NULL, time_int = NULL, ...)

Arguments

dat	[data.frame()] data frame of a pollyvote object.
na_handle	[character(1)] character describing which method is used to fill missing values. Currently supported is 'last': Use the last available value. Fill all days between the earliest and latest day in the data. 'omit': For every day, agregate available data. 'mean_within' Use the available data of this subcomponent to fill missing values. 'mean_across' Create a mini-pollyvote prediction by aggregating all source_types using na_handle = 'last' and then aggregating the result using na_handle = 'last', which will then be used to predict missing values in the lowest aggregation level.
pv	[pollyvote] only required for na_handle = 'mean_across', the pollyvote object containing the data frame.
...	[list()] additional arguments, currently ignored.

Value

the data set with filled mssing values.

pollyvote/pollyvoter documentation built on May 25, 2019, 11:23 a.m.