You may use the xout
argument in approx
"xout: an optional set of numeric values specifying where interpolation is to take place.
".
# create some fake data, which I _think_ may resemble the data you described in edit.
set.seed(123)
# "for t1 we have 1,3,5,7,9"
df1 <- data.frame(time = c(1, 3, 5, 7, 9), value = sample(1:10, 5))
df1
# "for t2 we have 1,2,3,4,5,6,7,8,9,10", the 'full time series'.
df2 <- data.frame(time = 1:10, value = sample(1:10))
# interpolate using approx and the xout argument
# The time values for 'full time series', df2$time, is used as `xout`.
# default values of arguments (e.g. linear interpolation, no extrapolation)
interpol1 <- with(df1, approx(x = time, y = value, xout = df2$time))
# some arguments you may wish to check
# extrapolation rules
interpol2 <- with(df1, approx(x = time, y = value, xout = df2$time,
rule = 2))
# interpolation method ('last observation carried forward")
interpol3 <- with(df1, approx(x = time, y = value, xout = df2$time,
rule = 2, method = "constant"))
df1
# time value
# 1 1 3
# 2 3 8
# 3 5 4
# 4 7 7
# 5 9 6
interpol1
# $x
# [1] 1 2 3 4 5 6 7 8 9 10
#
# $y
# [1] 3.0 5.5 8.0 6.0 4.0 5.5 7.0 6.5 6.0 NA
interpol3
# $x
# [1] 1 2 3 4 5 6 7 8 9 10
#
# $y
# [1] 3 3 8 8 4 4 7 7 6 6
# correlation between a vector of inter-(extra-)polated values
# and the 'full' time series
cor.test(interpol3$y, df2$value)