Hopach: distancematrix and distancevector do not return identical results
0
0
Entering edit mode
@01ec307d
Last seen 3 months ago
Netherlands

Hi,

the documentation of hopach says that the result of the distancevector and distancematrix functions should be "identical". However, that is not the case.

This is the example in the documentation:

mydata<-matrix(rnorm(50),nrow=10)
deuclid<-distancematrix(mydata,d="euclid")
# old method vdeuclid<-dissvector(deuclid)
vdeuclid<-deuclid@Data
ddaisy<-daisy(mydata)
vdeuclid ddaisy/sqrt(length(mydata[1,]))

d1<-distancematrix(mydata,d="abscosangle")
d2<-distancevector(mydata,mydata[1,],d="abscosangle")
d1[1,] d2 #equal to d1[1,]

However, at least since hopach-2.50, the results of distancevector apply also the extra step used in "ddaisy/sqrt(length(mydata[1,]))"

> mydata<-matrix(rnorm(50),nrow=10)
> deuclid<-distancematrix(mydata,d="euclid")
> vdeuclid<-deuclid@Data
> ddaisy<-daisy(mydata)
> vdeuclid[1:10]
[1] 3.534428 2.764194 4.274093 4.037899 1.942037 5.592038 2.539513 1.451171 4.565247 1.354340
> ddaisy/sqrt(length(mydata[1,]))[1]
Dissimilarities :
1         2         3         4         5         6         7         8         9
2  1.5806443
3  1.2361851 0.6056791
4  1.9114325 0.7752572 0.9199244
5  1.8058033 1.3887343 1.4681481 1.3347749
6  0.8685054 1.2025777 0.9410266 1.7518570 1.6624963
7  2.5008355 2.0364315 2.2338201 2.4665746 1.8328549 1.8465481
8  1.1357047 0.9515974 0.8468085 1.1752720 0.8702290 0.9018104 1.7221714
9  0.6489835 1.4214805 0.9261624 1.6151358 1.7102315 0.8523305 2.4401483 0.9605887
10 2.0416405 2.1729064 2.0948553 2.2511179 1.4792959 1.8247131 1.5366251 1.3427259 1.8488145

Metric :  euclidean
Number of objects : 10
> d1<-distancematrix(mydata,d="euclid")
> d2<-distancevector(mydata,mydata[1,],d="euclid")
> d1[1,]
1        2        3        4        5        6        7        8        9       10
1 0 3.534428 2.764194 4.274093 4.037899 1.942037 5.592038 2.539513 1.451171 4.565247
> d2
[1] 0.0000000 1.5806443 1.2361851 1.9114325 1.8058033 0.8685054 2.5008355 1.1357047 0.6489835 2.0416405
> d1[1,]/sqrt(length(mydata[1,]))
1        2        3        4        5         6        7        8         9       10
1 0 1.580644 1.236185 1.911433 1.805803 0.8685054 2.500836 1.135705 0.6489835 2.041641


Is this the intended behavior? Shouldn't both distancevector and distancematrix return the same without adding an extra step?? Since when is this happening?

Thanks,

Txema

> sessionInfo()
R version 4.1.2 (2021-11-01)
Platform: x86_64-apple-darwin17.0 (64-bit)
Running under: macOS Mojave 10.14.6

Matrix products: default
BLAS:   /System/Library/Frameworks/Accelerate.framework/Versions/A/Frameworks/vecLib.framework/Versions/A/libBLAS.dylib
LAPACK: /Library/Frameworks/R.framework/Versions/4.1/Resources/lib/libRlapack.dylib

Random number generation:
RNG:     Mersenne-Twister
Normal:  Inversion
Sample:  Rounding

locale:
[1] en_US.UTF-8/en_US.UTF-8/en_US.UTF-8/C/en_US.UTF-8/en_US.UTF-8

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

other attached packages:
[1] hopach_2.54.0       Biobase_2.54.0      BiocGenerics_0.40.0 cluster_2.1.2       sfsmisc_1.1-12
[6] pheatmap_1.0.12     gridExtra_2.3       ggplot2_3.3.5       SeuratObject_4.0.4  Seurat_4.1.0
[11] MASS_7.3-55         Matrix_1.4-0

loaded via a namespace (and not attached):
[1] nlme_3.1-155          spatstat.sparse_2.1-0 matrixStats_0.61.0    RcppAnnoy_0.0.19      RColorBrewer_1.1-2
[6] httr_1.4.2            sctransform_0.3.3     tools_4.1.2           utf8_1.2.2            R6_2.5.1
[11] irlba_2.3.5           rpart_4.1.16          KernSmooth_2.23-20    uwot_0.1.11           mgcv_1.8-38
[16] lazyeval_0.2.2        colorspace_2.0-2      withr_2.4.3           tidyselect_1.1.1      compiler_4.1.2
[21] cli_3.2.0             plotly_4.10.0         scales_1.1.1          spatstat.data_2.1-2   lmtest_0.9-39
[26] ggridges_0.5.3        pbapply_1.5-0         goftest_1.2-3         stringr_1.4.0         digest_0.6.29
[31] spatstat.utils_2.3-0  pkgconfig_2.0.3       htmltools_0.5.2       parallelly_1.30.0     fastmap_1.1.0
[36] htmlwidgets_1.5.4     rlang_1.0.1           shiny_1.7.1           generics_0.1.2        zoo_1.8-9
[41] jsonlite_1.7.3        spatstat.random_2.1-0 ica_1.0-2             dplyr_1.0.8           magrittr_2.0.2
[46] patchwork_1.1.1       Rcpp_1.0.8            munsell_0.5.0         fansi_1.0.2           abind_1.4-5
[51] reticulate_1.24       lifecycle_1.0.1       stringi_1.7.6         Rtsne_0.15            plyr_1.8.6
[56] grid_4.1.2            parallel_4.1.2        listenv_0.8.0         promises_1.2.0.1      ggrepel_0.9.1
[61] crayon_1.5.0          deldir_1.0-6          miniUI_0.1.1.1        lattice_0.20-45       cowplot_1.1.1
[66] splines_4.1.2         tensor_1.5            pillar_1.7.0          igraph_1.2.11         spatstat.geom_2.3-2
[71] future.apply_1.8.1    reshape2_1.4.4        codetools_0.2-18      leiden_0.3.9          glue_1.6.1
[76] data.table_1.14.2     BiocManager_1.30.16   png_0.1-7             vctrs_0.3.8           httpuv_1.6.5
[81] polyclip_1.10-0       gtable_0.3.0          RANN_2.6.1            purrr_0.3.4           spatstat.core_2.4-0
[86] tidyr_1.2.0           scattermore_0.8       future_1.23.0         mime_0.12             xtable_1.8-4
[91] later_1.3.0           survival_3.2-13       viridisLite_0.4.0     tibble_3.1.6          globals_0.14.0
[96] fitdistrplus_1.1-6    ellipsis_0.3.2        ROCR_1.0-11

documentation bug hopach euclidean distance • 162 views
ADD COMMENT

Login before adding your answer.

Traffic: 426 users visited in the last hour
Help About
FAQ
Access RSS
API
Stats

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.

Powered by the version 2.3.6