CODEWITHSUNDEEP
rstatisticsplyrapplylinear-regression
Kabau
Kabau

Reputation: 79

R: Selecting every two consecutive rows for ddplyr

This is my data

 Assay Sample Dilution  meanresp number
    1    S     0.25       68.55      1
    1    S     0.50       54.35      2
    1    S     1.00       44.75      3

My end goal is to apply a linear regression to every two consecutive rows and return the slope of that regression using Dilution and meanresp.

The length of the table can vary and i'd prefer not to use for loops as i'm trying to get out of the habit.

I think ddply would be good, but i'm not sure how to select the subset of every two consecutive rows. I thought perhaps there could be a way of saying 'do this for every vector subset of Dilution of length 2?

Any insight would be helpful.

Upvotes: 2

Views: 885

Answers (1)

LyzandeR
LyzandeR

Reputation: 37889

I don't know how this will be helpful in a linear regression but you could do something like that:

df <- read.table(header=T, text="Assay Sample Dilution  meanresp number
    1    S     0.25       68.55      1
    1    S     0.50       54.35      2
    1    S     1.00       44.75      3")

Using lapply:

> lapply(2:nrow(df), function(x) df[(x-1):x,] )
[[1]]
  Assay Sample Dilution meanresp number
1     1      S     0.25    68.55      1
2     1      S     0.50    54.35      2

[[2]]
  Assay Sample Dilution meanresp number
2     1      S      0.5    54.35      2
3     1      S      1.0    44.75      3

In case you want specific columns as well for your consecutive rows you can select them as:

> lapply(2:nrow(df), function(x) df[(x-1):x, c('Dilution','meanresp')] )
[[1]]
  Dilution meanresp
1     0.25    68.55
2     0.50    54.35

[[2]]
  Dilution meanresp
2      0.5    54.35
3      1.0    44.75

EDIT

If you want to perform a linear regression another lapply is enough to do it:

a <- lapply(2:nrow(df), function(x) df[(x-1):x, c('Dilution','meanresp')] )

b <- lapply(a,function(x) lm(Dilution~meanresp,data=x))

>b
[[1]]

Call:
lm(formula = Dilution ~ meanresp, data = x)

Coefficients:
(Intercept)     meanresp  
    1.45687     -0.01761  


[[2]]

Call:
lm(formula = Dilution ~ meanresp, data = x)

Coefficients:
(Intercept)     meanresp  
    3.33073     -0.05208

Or if you just want the slope:

b <- lapply(a, function(x) {
                    d <- lm(Dilution~meanresp,data=x)
                    coefficients(summary(d))[2,1]
})

> b
[[1]]
[1] -0.01760563

[[2]]
[1] -0.05208333

Upvotes: 3

Related Questions