How do I calculate the dot product in R using dplyr?

Question

My data is structure as follows:

dput(head(MovementAnalysis,10))
structure(list(Name = c("Amber", "Amber", "Amber", "Amber", "Amber", 
"Jeff", "Jeff", "Jeff", "Jeff", "Jeff"), Sample = c(1, 2, 3, 4, 5, 1, 2, 
3, 4, 5), X = c(26.66, 26.66, 26.65, 26.64, 26.64, 26.47, 26.46, 26.45, 
26.43, 26.42), Y = c(-12.38, -12.37, -12.36, -12.36, -12.35, -12.23, 
-12.22, -12.22, -12.22, -12.22)), .Names = c("Name", "Sample", "X", "Y"), row.names = c(NA, 10L), class = "data.frame")

I wish to calculate angular displacement, from the dot product of consecutive movement vectors, and angular velocity or the rate of change in angular displacement. I have used the code below but this does not restart the analysis according to a change in MovementAnalysis$Name? Simply, I wish to calculate angular velocity across MovementAnalysis$Sample for each person.

i <- 1
sampleRate <- 2
k <- as.integer(sampleRate)
a_velocity <- matrix(NA, length(MovementAnalysis$X))

 if (i > k)
      {
        a <- c(x[i] - x[i-k], y[i] - y[i-k])
        b <- c(x[i+k] - x[i], y[i+k] - y[i])
        a_velocity[i] <- acos(sum(a * b) / (sqrt(sum(a * a)) * sqrt(sum(b * b)))) * (180 / pi)
  }

  i <- i+1
}

I would like to use dplyr to complete the above but am unsure of the code? I have tried the following:

   function(x) {
     i <- 2

     while(i < length(X) - k) {
       if (i > k) {
         a <- c(X[i] - X[i-k], Y[i] - Y[i-k])
         b <- c(X[i+k] - X[i], Y[i+k] - Y[i])
         AngularVelocity = acos(sum(a * b) / (sqrt(sum(a * a)) * sqrt(sum(b * b)))) * (180 / pi)
       }

       i <- i+1
     }
   }

   MovementAnalysis <- MovementAnalysis %>%
     arrange(Name) %>%
     group_by(Name) %>%
     mutate(AngularV = function(x))

This returns an error of Error: unexpected ')' in: " group_by(Name) %>% mutate(AngularV = function(x))"Any ideas on the problem? The first 2 rows of Angular Velocity, for the first two rows of Name should be NA.

  SessionInfo()
  R version 3.1.2 (2014-10-31)
  Platform: i386-w64-mingw32/i386 (32-bit)

  locale:
  [1] LC_COLLATE=English_Australia.1252  LC_CTYPE=English_Australia.1252         
  LC_MONETARY=English_Australia.1252
  [4] LC_NUMERIC=C                       LC_TIME=English_Australia.1252    

  attached base packages:
  [1] stats     graphics  grDevices utils     datasets  methods   base     

  other attached packages:
  [1] dplyr_0.3.0.2 plyr_1.8.1    ggplot2_1.0.0

  loaded via a namespace (and not attached):
   [1] assertthat_0.1   colorspace_1.2-4 DBI_0.3.1        digest_0.6.4      
   grid_3.1.2       gtable_0.1.2    
   [7] labeling_0.3     lazyeval_0.1.9   magrittr_1.0.1   MASS_7.3-35      
   munsell_0.4.2    parallel_3.1.2  
   [13] proto_0.3-10     Rcpp_0.11.1      reshape2_1.4     scales_0.2.4     
   stringr_0.6.2    tools_3.1.2`

Your indentation made things hard to follow, I fixed it. 2-spaces per indent level works best for SO. — smci, Nov 14 '16 at 08:03

score 2 · Answer 1 · edited May 23 '17 at 12:17

2

First, don't use element-wise multiplication followed by sum: sum(a*b). Use the matrix-multiplication operator: a %*% b . So that would be: (a %*% b) / (sqrt(a %*% a) * sqrt(b %*% b))
But you're really trying to get cosine distance, which is a duplicate of: Find cosine similarity between two arrays

edited May 23 '17 at 12:17

Community

1
1

answered Nov 14 '16 at 08:02

smci

32,567
20
113
146

How do I calculate the dot product in R using dplyr?

1 Answers1