Heatmap

library(segtools)
library(dplyr)
library(ggplot2)

Motivating problem/issue

In earlier versions of the application, the heat-map background wasn’t smoothed like the original Excel application. This file documents how I changed the SEG graph using a pre-made .png image.

Data

To create the SEG using ggplot, I need to load a few data inputs from Github.. I can do this with the segtools::get_seg_data() function in the code chunk below:

risk_pairs -> this assigns a risk factor value to each BGM measurement
samp_measured_data -> this is a small sample measurement dataset
vand_data -> this is a dataset from Vanderbilt used to create some of the initial calculations

segtools::get_seg_data("names")
#>  [1] "VanderbiltComplete"   "AppRiskPairData"      "RiskPairData"        
#>  [4] "AppLookUpRiskCat"     "LookUpRiskCat"        "AppTestData"         
#>  [7] "AppTestDataSmall"     "AppTestDataMed"       "AppTestDataBig"      
#> [10] "FullSampleData"       "ModBAData"            "No_Interference_Dogs"
#> [13] "SEGRiskTable"         "SampMeasData"         "SampleData"          
#> [16] "lkpRiskGrade"         "lkpSEGRiskCat4"

risk_pairs <- segtools::get_seg_data("AppRiskPairData")
samp_measured_data <- segtools::get_seg_data("FullSampleData")
vand_data <- segtools::get_seg_data("VanderbiltComplete")

mmolConvFactor <- segtools::mmolConvFactor
mmolConvFactor
#> [1] 18.01806

risk_factor_colors <- segtools::risk_factor_colors

base_data <- segtools::base_data
base_data
#>   x_coordinate y_coordinate color_gradient
#> 1            0            0              0
#> 2            0            0              1
#> 3            0            0              2
#> 4            0            0              3
#> 5            0            0              4

risk_pairs has columns and risk pairs for both REF and BGM, and the RiskFactor variable for each pair of REF and BGM data. Below you can see a sample of the REF, BGM, RiskFactor, and abs_risk variables.

risk_pairs |> 
  dplyr::sample_n(size = 10) |>  
  dplyr::select(
    REF, BGM, RiskFactor, abs_risk)
#> # A tibble: 10 × 4
#>      REF   BGM RiskFactor abs_risk
#>    <dbl> <dbl>      <dbl>    <dbl>
#>  1   396   490     -0.366    0.366
#>  2    79   558     -3.18     3.18 
#>  3   594    42      3.12     3.12 
#>  4   527   215      1.11     1.11 
#>  5   333   570     -0.697    0.697
#>  6   177   539     -2.04     2.04 
#>  7   315   378     -0.211    0.211
#>  8   591   434      0.155    0.155
#>  9   203   144      0.919    0.919
#> 10   474   335      0.514    0.514

samp_measured_data mimics a blood glucose monitor, with only BGM and REF values.

samp_measured_data |> 
  dplyr::sample_n(size = 10) |>  
  dplyr::select(
    REF, BGM)
#> # A tibble: 10 × 2
#>      REF   BGM
#>    <dbl> <dbl>
#>  1   121   123
#>  2   115   120
#>  3   111   116
#>  4   132   143
#>  5   194   202
#>  6   179   174
#>  7   150   161
#>  8   109   100
#>  9   174   183
#> 10   106   111

vand_data contains blood glucose monitor measurements, with only BGM and REF values.

vand_data |>  
  dplyr::sample_n(size = 10) |> 
  dplyr::select(REF, BGM)
#> # A tibble: 10 × 2
#>      REF   BGM
#>    <dbl> <dbl>
#>  1   264   257
#>  2   122   129
#>  3   116   108
#>  4   136   126
#>  5   131   131
#>  6   124   132
#>  7   158   153
#>  8   101   106
#>  9   158   154
#> 10   132   143

The original (Excel) SEG image

Below is the image from the Excel file:

The points are plotted against a Gaussian smoothed background image.

`ggplot2` image using risk pairs

The steps/code to create the current ggplot2 image are below

Base layer

ggp_base <- ggplot() +
  geom_point(data = base_data, 
    aes(x = x_coordinate,
      y = y_coordinate,
      fill = color_gradient))
ggp_base

Risk color gradient

The risk layer adds the risk_pairs data and creates the color gradient.

ggp_risk_color_gradient <- ggp_base +
  geom_point(data = risk_pairs, 
    aes(x = REF, 
        y = BGM,
        color = abs_risk), show.legend = FALSE)  +
  ggplot2::scale_color_gradientn(
    colors = risk_factor_colors, 
    guide = "none",
    limits = c(0, 4),
    values = scales::rescale(c(
      0, # darkgreen
      0.4375, # green
      1.0625, # yellow
      2.7500, # red
      4.0000)))
ggp_risk_color_gradient

Risk fill gradient

The guide is added with ggplot2::scale_fill_gradientn() and ggplot2::guide_colorbar().

ggp2_risk_fill_gradient <- ggp_risk_color_gradient +
  ggplot2::scale_fill_gradientn(values = scales::rescale(c(
      0, # darkgreen
      0.4375, # green
      1.0625, # yellow
      2.75, # red
      4.0 # brown
    )), 
    limits = c(0, 4),
    colors = risk_factor_colors,
    guide = ggplot2::guide_colorbar(ticks = FALSE,
                           barheight = unit(100, "mm")),
                           breaks = c(0.25, 1, 2, 3, 3.75),
                           labels = c("None", "Slight", 
                                      "Moderate", "High", "Extreme"),
                           name = "Risk level")
ggp2_risk_fill_gradient

Scales

The x and y scales are set manually using ggplot2::scale_y_continuous() and ggplot2::scale_x_continuous().

ggp_scales <- ggp2_risk_fill_gradient +
  ggplot2::scale_y_continuous(
    limits = c(0, 600),
    sec.axis =
      sec_axis(~. / mmolConvFactor,
        name = "Measured blood glucose (mmol/L)"
      ),
    name = "Measured blood glucose (mg/dL)"
  ) +
  ggplot2::scale_x_continuous(
    limits = c(0, 600),
    sec.axis =
      sec_axis(~. / mmolConvFactor,
        name = "Reference blood glucose (mmol/L)"
      ),
    name = "Reference blood glucose (mg/dL)"
  )
ggp_scales

Theme

Finally, the theme is added to polish the output.

ggp_seg <- ggp_scales + 
  segtools::theme_seg(base_size = 20)
ggp_seg