SciencesPo (version 1.02.12)

tsCollapse: Aggregated Data by Time Parameters

Description

Return a data.frame with aggregated data by time parameters. In order to use this function, you have to have a as.timedf data.frame.

Usage

tsCollapse(data, by = c("day", "month", "year"), FUN, na.rm = FALSE,
  factor = NULL, civil = FALSE, plot = TRUE)

Arguments

data

a timedf object

by

the method to break up the dependent variable; options: "day","month", or "year".

FUN

a function to aggregate data sum, mean, min, max, etc.

na.rm

A logical value for na.rm, default is FALSE

factor

is the aggregator factor; a numeric value representing days, months or years to perform aggregation by.

civil

Wether the years start from the civil date or from the first observation in data. The default is civil = FALSE.

plot

whether the data.frame to be plotted, default is plot = TRUE.

encoding

UTF-8

Details

A data.frame of type long with summarized time series data (y) and time parameters formatted as POSIXct.

Examples

data(us2012)
Obama.ts <- as.timedf(us2012[,3], '%Y-%m-%d', us2012[,8])

# Daily aggregated means for data:
daily <- tsCollapse(Obama.ts, by = "day", mean)

# Weekly aggregated means for data:
weekly <- tsCollapse(Obama.ts, by = "day", factor = 7, mean)

# monthly aggregated means for data:
monthly <- tsCollapse(Obama.ts, by = "month", mean)

# bimonthly or semimonthly aggregated means for data:
bimonthly <- tsCollapse(Obama.ts,by = "month", mean, factor = 2)