Skip to content

Latest commit

 

History

History
426 lines (313 loc) · 6.17 KB

Types_of_Distances_in_R_Python.md

File metadata and controls

426 lines (313 loc) · 6.17 KB
Raw

distance-similarity

index



preparation for using R and python

library (reticulate)
path <- Sys.which("python")
path <- gsub("\\", "//", path, fixed = TRUE)
use_python(path)



Eculidean Distance Using R (i)

set.seed(123)
Mat <- matrix(rnorm(40, 2, 2) |> round(2), 2, 20)
head(Mat)
     [,1] [,2] [,3]  [,4] [,5] [,6] [,7] [,8]  [,9] [,10] [,11] [,12] [,13]
[1,] 0.88 5.12 2.26  2.92 0.63 4.45 2.80 0.89  3.00  3.40 -0.14 -0.05  0.75
[2,] 1.54 2.14 5.43 -0.53 1.11 2.72 2.22 5.57 -1.93  1.05  1.56  0.54 -1.37
     [,14] [,15] [,16] [,17] [,18] [,19] [,20]
[1,]  3.68 -0.28  2.85  3.79  3.64  3.11  1.39
[2,]  2.31  4.51  1.41  3.76  3.38  1.88  1.24

dist(Mat, method = "euclidean")
         1
2 11.08097



Eculidean Distance Using python (i)

from scipy.spatial import distance

# Calculate Euclidean distance
dist = distance.euclidean(r.Mat[0, :], r.Mat[1, :])
print(round(dist, 5))
11.08097




Minkovski Distance Using R (ii)

k <- 5
dist(Mat, method = "minkowski", p = k)
         1
2 6.161605



Minkovski Distance Using Python (ii)

distance.minkowski(r.Mat[0, :], r.Mat[1, :], r.k)
6.161604866392561




Manhattan Distance Using R (iii)

dist(Mat, method = "manhattan")
      1
2 38.69



Manhattan Distance Using Python (iii)

distance.cityblock(r.Mat[0, :], r.Mat[1, :])
38.68999999999999




hamming Distance Using R (iv)

x <- c(1, -1, 1, 1, -1, 1, 1)
y <- c(-1, -1, 1, 1, 1, -1, 1)
Hamming_dist <- sum(x != y)
Hamming_dist
[1] 3



hamming Distance Using Python (iv)

distance.hamming(r.x, r.y) * len(r.x)
3.0




Cosine Distance Using R (v)

if(!require(stylo)) {
    chooseCRANmirror(graphics = F, ind = 1)
    install.packages("stylo")
}
Loading required package: stylo


### stylo version: 0.7.4 ###

If you plan to cite this software (please do!), use the following reference:
    Eder, M., Rybicki, J. and Kestemont, M. (2016). Stylometry with R:
    a package for computational text analysis. R Journal 8(1): 107-121.
    <https://journal.r-project.org/archive/2016/RJ-2016-007/index.html>

To get full BibTeX entry, type: citation("stylo")

x <- c(3, 4, 1, 3, 5, 10)
y <- c(1, 0, 3, 5, 10, 11)
cosine_dist <- stylo :: dist.cosine(rbind(x, y))
cosine_dist
          x
y 0.1056683



Cosine Distance Using Python (v)

distance.cosine(r.x, r.y) 
0.10566834923363022




gower Index Using R (vi)

if(!require(cluster)) {
    chooseCRANmirror(graphics = F, ind = 1)
    install.packages("cluster")
}
Loading required package: cluster

df <- data.frame(
  num = c(1, 2, 3, 4),
  cat = factor(c("A", "B", "A", "B"))
)

df
  num cat
1   1   A
2   2   B
3   3   A
4   4   B

gower_dist <- cluster :: daisy(df, metric = "gower")

print(gower_dist)
Dissimilarities :
          1         2         3
2 0.6666667                    
3 0.3333333 0.6666667          
4 1.0000000 0.3333333 0.6666667

Metric :  mixed ;  Types = I, N 
Number of objects : 4



gower Index Using Python (vi)

import PyDistances as distt
import pandas as pd
distt.Gower_Similarity_Matrix(r.df, p1 = 1, p2 = 0, p3 = 1)
array([[1.        , 0.33333333, 0.66666667, 0.        ],
       [0.33333333, 1.        , 0.33333333, 0.66666667],
       [0.66666667, 0.33333333, 1.        , 0.33333333],
       [0.        , 0.66666667, 0.33333333, 1.        ]])




Czekanowski-Sorensen Index Using R (vii)

x <- c('a', 'a', 'b', 'c', 'b', 'b', 'a')
y <- c('b', 'd', 'a', 'b', 'a', 'b', 'c')

dist_sorensen <- 2 * (intersect(x, y) |> length()) / 
                    (x |> unique() |> length() + y |> unique()|> length()) 
dist_sorensen
[1] 0.8571429




Czekanowski-Sorensen Index Using python (vii)

2 * len(set(r.x).intersection(set(r.y))) / (len(set(r.x)) + len(set(r.y)))
0.8571428571428571




Jaccard Index Using R (viii)

x <- c('a', 'a', 'b', 'c', 'b', 'b', 'a')
y <- c('b', 'd', 'a', 'b', 'a', 'b', 'c')

dist_Jaccard <- (intersect(x, y) |> length()) / 
                    (union(x, y) |> length()) 
dist_Jaccard
[1] 0.75



Jaccard Index Using python (viii)

len(set(r.x).intersection(set(r.y))) / (len(set(r.x).union(set(r.y))))
0.75