Predict.interval: Make prediction intervals for the time series

In St-Digital-Twin/Dtwin: Re-usable functions for City Dtwin

Description

prepare high & low values for time series

Usage

Predict.interval (df, loc, ind, kpi, periods, interval.width)

Arguments

df	dataframe which have to include the next columns: scenario, year, value
loc	name of location where we create interval
ind	name of industry where we create interval
kpi	name of indicator which we create interval
periods	the number of months
interval.width	the size from 0.7 till 0.99

Value

dataframe

Examples

num_loc <- distinct(data_10, location) %>% arrange(location) %>% rowid_to_column() %>% mutate(total = nrow(.))
num_ind <- distinct(data_10, industry) %>% arrange(industry   ) %>% rowid_to_column() %>% mutate(total = nrow(.))
num_kpi <- distinct(data_10, indicator) %>% arrange(indicator) %>% rowid_to_column() %>% mutate(total = nrow(.))
data_1 <- data_10 %>%
 nest_by(location,indicator,industry) %>%
 mutate(data = list(Prediction_interval(data,
                                        loc = num_loc[grepl(location , num_loc$location  ),],
                                        ind = num_ind[grepl(industry , num_ind$industry  ),],
                                        kpi = num_kpi[grepl(indicator, num_kpi$indicator ),]))) %>%
 unnest(data) %>% ungroup() %>%
 mutate(year = as.integer(year),
        value = ifelse (value<0,0,value))

St-Digital-Twin/Dtwin documentation built on Jan. 1, 2022, 8:11 p.m.