How to create a correlation matrix between two different databases?

General
,
1

Hello everyone, looking for information on the internet about how to create a matrix correlation in R, I realized that in all the examples they compare the same variable to each other. In my case I seek to compare between two different databases, but that are related to each other.

I found an example in a article of what type of correlation matrix i want to get:

image
image1793×1566 411 KB

My databases are very similar to the example above.

metadata <- data.frame(tibble::tribble(
                          ~SampleID,  ~DO,   ~pH, ~Temperature, ~Turbidity,
                               "1A", 4.24,  9.94,        24.48,       87.3,
                               "1B", 2.58, 10.06,        24.95,       95.2,
                               "1C", 2.85,  9.98,        24.01,       98.8,
                               "1D", 3.81,  9.07,        26.67,       24.9,
                               "1E", 2.64,  8.99,        24.01,       43.1,
                               "1F", 4.61,  8.93,         23.5,       49.1,
                               "1G", 5.74,  9.24,        24.46,       68.6,
                               "1H", 3.69,  9.32,        24.65,       69.8,
                               "2A", 4.24,  9.94,        24.48,       87.3,
                               "2B", 2.66,    10,        24.92,        113,
                               "2C", 2.85,  9.98,        24.01,       98.8,
                               "2D", 3.81,  9.07,        26.67,       24.9,
                               "2E", 2.64,  8.99,        24.01,       43.1,
                               "2F", 4.61,  8.93,         23.5,       49.1,
                               "2G", 5.74,  9.24,        24.46,       68.6,
                               "2H", 3.69,  9.32,        24.65,       69.8
                          )
)

bacteria <- data.frame (tibble::tribble(
                           ~index, ~Un.Melainabacteria, ~Nostocaceae, ~Un.Nostocales, ~Phormidesmiaceae, ~Un.Phormidesmiales, ~Cyanobiaceae, ~Un.Oxyphotobacteria, ~Un.Cyanobacteria,
                             "1A",                   0,            0,             12,                66,                   0,             0,                 7137,              1467,
                             "1B",                   0,            0,              6,                37,                   0,             0,                 5549,               238,
                             "1C",                   0,           19,              0,                96,                   0,             0,                 1861,               708,
                             "1D",                   0,            0,              0,                66,                   0,            19,                 3185,              4001,
                             "1E",                 273,         1662,            207,               720,                  67,           529,                 7468,              1928,
                             "1F",                 470,         2090,            425,               563,                  30,            23,                18421,              4460,
                             "1G",                   0,          372,              0,                19,                   0,             2,                 4191,               361,
                             "1H",                   0,          433,             46,              1261,                  33,             0,                12240,              1970,
                             "2A",                 363,         1000,              8,               195,                  37,             2,                20728,              4350,
                             "2B",                 110,          769,             62,                46,                 108,           648,                13438,              3076,
                             "2C",                 169,          639,             31,                14,                  11,            21,                17664,              2008,
                             "2D",                   0,          510,              0,               118,                  14,             0,                 5085,              2745,
                             "2E",                 129,          319,              0,                60,                   0,           200,                 8376,              1613,
                             "2F",                   0,          501,             89,               818,                   0,             0,                 8906,               983,
                             "2G",                   9,          930,            115,               167,                  16,           195,                13273,              1514,
                             "2H",                 209,         1876,            535,               316,                  14,            43,                18506,              3531
                           )
)

Created on 2019-10-31 by the reprex package (v0.3.0)

2

Hi @OSDIAZ. You can create the correlation matrix first and filter the matrix. Finally, use corrplot function from corrplot package with argument is.corr = FALSE for non square matrix.

library(tidyverse)

metadata <- data.frame(tibble::tribble(
  ~SampleID,  ~DO,   ~pH, ~Temperature, ~Turbidity,
  "1A", 4.24,  9.94,        24.48,       87.3,
  "1B", 2.58, 10.06,        24.95,       95.2,
  "1C", 2.85,  9.98,        24.01,       98.8,
  "1D", 3.81,  9.07,        26.67,       24.9,
  "1E", 2.64,  8.99,        24.01,       43.1,
  "1F", 4.61,  8.93,         23.5,       49.1,
  "1G", 5.74,  9.24,        24.46,       68.6,
  "1H", 3.69,  9.32,        24.65,       69.8,
  "2A", 4.24,  9.94,        24.48,       87.3,
  "2B", 2.66,    10,        24.92,        113,
  "2C", 2.85,  9.98,        24.01,       98.8,
  "2D", 3.81,  9.07,        26.67,       24.9,
  "2E", 2.64,  8.99,        24.01,       43.1,
  "2F", 4.61,  8.93,         23.5,       49.1,
  "2G", 5.74,  9.24,        24.46,       68.6,
  "2H", 3.69,  9.32,        24.65,       69.8
)
)

bacteria <- data.frame (tibble::tribble(
  ~index, ~Un.Melainabacteria, ~Nostocaceae, ~Un.Nostocales, ~Phormidesmiaceae, ~Un.Phormidesmiales, ~Cyanobiaceae, ~Un.Oxyphotobacteria, ~Un.Cyanobacteria,
  "1A",                   0,            0,             12,                66,                   0,             0,                 7137,              1467,
  "1B",                   0,            0,              6,                37,                   0,             0,                 5549,               238,
  "1C",                   0,           19,              0,                96,                   0,             0,                 1861,               708,
  "1D",                   0,            0,              0,                66,                   0,            19,                 3185,              4001,
  "1E",                 273,         1662,            207,               720,                  67,           529,                 7468,              1928,
  "1F",                 470,         2090,            425,               563,                  30,            23,                18421,              4460,
  "1G",                   0,          372,              0,                19,                   0,             2,                 4191,               361,
  "1H",                   0,          433,             46,              1261,                  33,             0,                12240,              1970,
  "2A",                 363,         1000,              8,               195,                  37,             2,                20728,              4350,
  "2B",                 110,          769,             62,                46,                 108,           648,                13438,              3076,
  "2C",                 169,          639,             31,                14,                  11,            21,                17664,              2008,
  "2D",                   0,          510,              0,               118,                  14,             0,                 5085,              2745,
  "2E",                 129,          319,              0,                60,                   0,           200,                 8376,              1613,
  "2F",                   0,          501,             89,               818,                   0,             0,                 8906,               983,
  "2G",                   9,          930,            115,               167,                  16,           195,                13273,              1514,
  "2H",                 209,         1876,            535,               316,                  14,            43,                18506,              3531
)
)

corrr::correlate(bind_cols(metadata[, -1], bacteria[, -1])) %>%
  filter(rowname %in% colnames(metadata)) %>%
  select(one_of(c("rowname", colnames(bacteria)[-1]))) %>%
  as.data.frame() %>%
  column_to_rownames("rowname") %>%
  as.matrix() %>%
  corrplot::corrplot(is.corr = FALSE)
#> 
#> Correlation method: 'pearson'
#> Missing treated using: 'pairwise.complete.obs'

Created on 2019-10-31 by the reprex package (v0.3.0)

6 Likes
3

This topic was automatically closed 7 days after the last reply. New replies are no longer allowed.