Two R functions for finding dry spells in precipitation time series.
This script was written to filter out dry spells in climate time series data and obtain a human-readable summary.
count_consecutive(x) retains a numeric vector with a count of days since the last precipitation was measured.
dryspells_readable(x, dates, threshold = 0) creates a dataframe with start and end date of dryspells, and gives the length in a separate column.
Before you can use the function, make sure to source("count_consecutive.R") resp. source("dryspells_readable.R") to load the function into your environment. Alternatively, you can also copy the function code into your own script and parse it before using.
## For a simple count since the last precipitation measured
count_consecutive(x)
## For a readable start and end dataframe with length of the period
dryspells_readable(x, dates, threshold = 0)
- x
- A numeric vector, containing the measured precipitation.
- dates
- A vector of dates, formatted as a date class. Will be used for the output of start and end dates as well as for calculating the length of the period.
- threshold
- Numeric or integer stating up to which amount of measured precipitation a day is considered to be dry.
# create random data for testing
measurements <- data.frame(
precipitation = abs(rnorm(200, 0, 1)),
date = seq.Date(
from = as.Date("2017-05-01"),
by = "day",
length.out = 200
)
)
# to obtain some more dry periods in the test data
measurements$precipitation[
measurements$precipitation < .5
] <- 0
# source functions
source("count_consecutive.R")
source("dryspells_readable.R")
# run functions
dry_daycount <- count_consecutive(measurements$precipitation)
dry_readable <- find_dryspells(
x = measurements$precipitation,
dates = measurements$date,
threshold = .6
)
# view the results
print(dry_daycount)
View(dry_readable)