Graph with multiple variables

General
1

I am trying to plot the below data which contains 683 samples. I cannot find a suitable graph that can plot all the data at once to get insights between the variables. A part from doing individual graph for each variables is there any other ways anyone can suggest please?

CancerData=data.frame(stringsAsFactors=FALSE,
                Id = c("1000025", "1002945", "1015425", "1016277", "1017023"),
      Cl.thickness = c("5", "5", "3", "6", "4"),
         Cell.size = c("1", "4", "1", "8", "1"),
        Cell.shape = c("1", "4", "1", "8", "1"),
     Marg.adhesion = c("1", "5", "1", "NA", "3"),
      Epith.c.size = c("2", "7", "2", "3", "2"),
       Bare.nuclei = as.factor(c("1", "10", "2", "4", "1")),
       Bl.cromatin = as.factor(c("3", "3", "3", "3", "3")),
   Normal.nucleoli = as.factor(c("1", "2", "1", "7", "1")),
           Mitoses = as.factor(c("1", "NA", "1", "1", "1")),
             Class = as.factor(c("benign", "benign", "benign", "benign",
                                 "benign"))
)


plot(CancerData)
#> Warning in data.matrix(x): NAs introduced by coercion

Created on 2019-11-21 by the reprex package (v0.3.0)

2

Hi @user124578. You may try ggpairs function from GGally like the following code. The plot from your sample data seems not attractive because your sample data contains all character or factor data. You can visit Extension to ggplot2 • GGally for tutorials.

library(tidyverse)

CancerData=data.frame(stringsAsFactors=FALSE,
                     Id = c("1000025", "1002945", "1015425", "1016277", "1017023"),
                     Cl.thickness = c("5", "5", "3", "6", "4"),
                     Cell.size = c("1", "4", "1", "8", "1"),
                     Cell.shape = c("1", "4", "1", "8", "1"),
                     Marg.adhesion = c("1", "5", "1", "NA", "3"),
                     Epith.c.size = c("2", "7", "2", "3", "2"),
                     Bare.nuclei = as.factor(c("1", "10", "2", "4", "1")),
                     Bl.cromatin = as.factor(c("3", "3", "3", "3", "3")),
                     Normal.nucleoli = as.factor(c("1", "2", "1", "7", "1")),
                     Mitoses = as.factor(c("1", "NA", "1", "1", "1")),
                     Class = as.factor(c("benign", "benign", "benign", "benign",
                                         "benign"))
)

CancerData %>%
 column_to_rownames("Id") %>%
 GGally::ggpairs()
#> Registered S3 method overwritten by 'GGally':
#>   method from   
#>   +.gg   ggplot2

Created on 2019-11-21 by the reprex package (v0.3.0)

3

I think you need to be more specific about what kind of insights are you looking for, for example, if you want to see how variables correlate with each other you could use a correlation matrix.

library(corrplot)
library(tidyverse)

CancerData <- data.frame(stringsAsFactors=FALSE,
                         Id = c("1000025", "1002945", "1015425", "1016277", "1017023"),
                         Cl.thickness = c("5", "5", "3", "6", "4"),
                         Cell.size = c("1", "4", "1", "8", "1"),
                         Cell.shape = c("1", "4", "1", "8", "1"),
                         Marg.adhesion = c("1", "5", "1", "NA", "3"),
                         Epith.c.size = c("2", "7", "2", "3", "2"),
                         Bare.nuclei = as.factor(c("1", "10", "2", "4", "1")),
                         Bl.cromatin = as.factor(c("3", "3", "3", "3", "3")),
                         Normal.nucleoli = as.factor(c("1", "2", "1", "7", "1")),
                         Mitoses = as.factor(c("1", "NA", "1", "1", "1")),
                         Class = as.factor(c("benign", "benign", "benign", "benign",
                                             "benign"))
)

CancerData <- CancerData %>% 
    mutate_at(vars(-Class), ~parse_number(as.character(.)))

CancerData %>% 
    select(-Id, -Class) %>% 
    cor() %>% 
    corrplot(method = "circle")

2 Likes
4

I am trying to find out how each variable affected with the class type.

5

This topic was automatically closed 21 days after the last reply. New replies are no longer allowed.