CODEWITHSUNDEEP
r
dom
dom

Reputation: 29

R: add rows in dataframe depending on columnvalue

I have a data frame including a motion signal sampled at 100Hz. In a simplified way it looks like this:

df<-data.frame(c(0.01, 0.02, 0.05, 0.06), c(0, 0, 2, 0), c(2,3,10,15))
names(df)<-c("counter", "lostSamples", "value")
df
#   counter lostSamples value
# 1    0.01           0     2
# 2    0.02           0     3
# 3    0.05           2    10
# 4    0.06           0    15

counter” is each sample; “lostSamples” is when my device could not record/store the sample; “value” is the actual signal in 1 dimension relative to a reference

I would like to add rows when there are lost samples. In this example I have to fill the gap (add rows to data frame) for counters 0.3 and 0.4. I would like to fill the gaps either with NAs (easier) or interpolating by a linear function depending on values before and after the gap.

I tried with for loops and with apply but was not successful.

Upvotes: 2

Views: 96

Answers (2)

josliber
josliber

Reputation: 44340

If you have the list of expected values for the counter, you can do this with the approx function for linear interpolation:

expected <- seq(0.01, 0.06, 0.01)
data.frame(counter=expected,
           value=approx(df$counter, df$value, expected)$y)
#   counter     value
# 1    0.01  2.000000
# 2    0.02  3.000000
# 3    0.03  5.333333
# 4    0.04  7.666667
# 5    0.05 10.000000
# 6    0.06 15.000000

Upvotes: 1

Veerendra Gadekar
Veerendra Gadekar

Reputation: 4472

This could be another way

df2 = data.frame(counter = seq(0.01, 0.06, 0.01))
out = merge(df, df2, all = T)

#  counter lostSamples value
#1    0.01           0     2
#2    0.02           0     3
#3    0.03          NA    NA
#4    0.04          NA    NA
#5    0.05           2    10
#6    0.06           0    15

Upvotes: 2

Related Questions