\([![DOI](https://zenodo.org/badge/DOI/10.5281/zenodo.3635430.svg)](https://doi.org/10.5281/zenodo.3635430)\)

https://doi.org/10.5281/zenodo.3635430

https://osf.io/3tjfk/

Histopathology Research Template 🔬


1 Introduction

  • State the marker of interest, the study objectives, and hypotheses (Knijn, Simmer, and Nagtegaal 2015).1

2 Materials & Methods

Describe Materials and Methods as highlighted in (Knijn, Simmer, and Nagtegaal 2015).2

  • Describe patient characteristics, and inclusion and exclusion criteria

  • Describe treatment details

  • Describe the type of material used

  • Specify how expression of the biomarker was assessed

  • Describe the number of independent (blinded) scorers and how they scored

  • State the method of case selection, study design, origin of the cases, and time frame

  • Describe the end of the follow-up period and median follow-up time

  • Define all clinical endpoints examined

  • Specify all applied statistical methods

  • Describe how interactions with other clinical/pathological factors were analyzed


2.1 Statistical Methods


2.2 Header Codes

Codes for general settings.3

Setup global chunk settings4

knitr::opts_chunk$set(
    eval = TRUE,
    echo = TRUE,
    fig.path = here::here("figs/"),
    message = FALSE,
    warning = FALSE,
    error = TRUE,
    cache = TRUE,
    comment = NA,
    tidy = TRUE,
    fig.width = 6,
    fig.height = 4
)
library(knitr)
hook_output = knit_hooks$get("output")
knit_hooks$set(output = function(x, options) {
    # this hook is used only when the linewidth option is not NULL
    if (!is.null(n <- options$linewidth)) {
        x = knitr:::split_lines(x)
        # any lines wider than n should be wrapped
        if (any(nchar(x) > n)) 
            x = strwrap(x, width = n)
        x = paste(x, collapse = "\n")
    }
    hook_output(x, options)
})
# linewidth css
  pre:not([class]) {
    color: #333333;
    background-color: #cccccc;
  }
# linewidth css
pre.jamovitable{
  color:black;
  background-color: white;
  margin-bottom: 35px;  
}
jtable <- function(jobject, digits = 3) {
    snames <- sapply(jobject$columns, function(a) a$title)
    asDF <- jobject$asDF
    tnames <- unlist(lapply(names(asDF), function(n) snames[[n]]))
    names(asDF) <- tnames
    kableExtra::kable(asDF, "html", table.attr = "class=\"jmv-results-table-table\"", 
        row.names = F, digits = 3)
}

Block rmdnote

Block rmdtip

Block warning

Load Library

see R/loadLibrary.R for the libraries loaded.

source(file = here::here("R", "loadLibrary.R"))

2.3 Generate Fake Data

Codes for generating fake data.5

Generate Fake Data

This code generates a fake histopathological data. Some sources for fake data generation here6 , here7 , here8 , here9 , here10 , here11 , here12 , here13 , and here14 .

Use this code to generate fake clinicopathologic data

source(file = here::here("R", "gc_fake_data.R"))
wakefield::table_heat(x = fakedata, palette = "Set1", flip = TRUE, print = TRUE)


2.4 Import Data

Codes for importing data.15

Read the data

library(readxl)
mydata <- readxl::read_excel(here::here("data", "mydata.xlsx"))
# View(mydata) # Use to view data after importing

Add code for import multiple data purrr reduce


2.5 Study Population

2.5.1 Report General Features

Codes for reporting general features.16

Dataframe Report

# Dataframe report
mydata %>% dplyr::select(-contains("Date")) %>% report::report(.)
The data contains 250 observations of the following variables:
  - ID: 250 entries: 001, n = 1; 002, n = 1; 003, n = 1 and 247 others (0 missing)
  - Name: 249 entries: Aceyn, n = 1; Adalaide, n = 1; Adidas, n = 1 and 246 others (1 missing)
  - Sex: 2 entries: Male, n = 127; Female, n = 122 (1 missing)
  - Age: Mean = 49.54, SD = 14.16, Median = , MAD = 17.79, range: [25, 73], Skewness = 0.00, Kurtosis = -1.15, 1 missing
  - Race: 7 entries: White, n = 158; Hispanic, n = 38; Black, n = 30 and 4 others (1 missing)
  - PreinvasiveComponent: 2 entries: Absent, n = 203; Present, n = 46 (1 missing)
  - LVI: 2 entries: Absent, n = 147; Present, n = 102 (1 missing)
  - PNI: 2 entries: Absent, n = 171; Present, n = 78 (1 missing)
  - Death: 2 levels: FALSE (n = 83, 33.20%); TRUE (n = 166, 66.40%) and missing (n = 1, 0.40%)
  - Group: 2 entries: Treatment, n = 131; Control, n = 118 (1 missing)
  - Grade: 3 entries: 3, n = 109; 1, n = 78; 2, n = 62 (1 missing)
  - TStage: 4 entries: 4, n = 118; 3, n = 65; 2, n = 43 and 1 other (0 missing)
  - AntiX_intensity: Mean = 2.39, SD = 0.66, Median = , MAD = 1.48, range: [1, 3], Skewness = -0.63, Kurtosis = -0.65, 1 missing
  - AntiY_intensity: Mean = 2.02, SD = 0.80, Median = , MAD = 1.48, range: [1, 3], Skewness = -0.03, Kurtosis = -1.42, 1 missing
  - LymphNodeMetastasis: 2 entries: Absent, n = 144; Present, n = 105 (1 missing)
  - Valid: 2 levels: FALSE (n = 116, 46.40%); TRUE (n = 133, 53.20%) and missing (n = 1, 0.40%)
  - Smoker: 2 levels: FALSE (n = 130, 52.00%); TRUE (n = 119, 47.60%) and missing (n = 1, 0.40%)
  - Grade_Level: 3 entries: high, n = 109; low, n = 77; moderate, n = 63 (1 missing)
  - DeathTime: 2 entries: Within1Year, n = 149; MoreThan1Year, n = 101 (0 missing)
mydata %>% explore::describe_tbl()
250 observations with 21 variables
18 variables containing missings (NA)
0 variables with no variance

2.6 Ethics and IRB

2.6.1 Always Respect Patient Privacy

Always Respect Patient Privacy
- Health Information Privacy17
- Kişisel Verilerin Korunması18


2.7 Define Variable Types

Codes for defining variable types.19

print column names as vector

dput(names(mydata))
c("ID", "Name", "Sex", "Age", "Race", "PreinvasiveComponent", 
"LVI", "PNI", "LastFollowUpDate", "Death", "Group", "Grade", 
"TStage", "AntiX_intensity", "AntiY_intensity", "LymphNodeMetastasis", 
"Valid", "Smoker", "Grade_Level", "SurgeryDate", "DeathTime")

2.7.1 Find Key Columns

2.7.1.1 Find ID and key columns to exclude from analysis

vctrs::vec_assert()

dplyr::all_equal()

arsenal::compare()

visdat::vis_compare()

See the code as function in R/find_key.R.

keycolumns <- mydata %>% sapply(., FUN = dataMaid::isKey) %>% tibble::as_tibble() %>% 
    dplyr::select(which(.[1, ] == TRUE)) %>% names()
keycolumns
[1] "ID"   "Name"

2.7.2 Variable Types

Get variable types

mydata %>% dplyr::select(-keycolumns) %>% inspectdf::inspect_types()
# A tibble: 4 x 4
  type             cnt  pcnt col_name  
  <chr>          <int> <dbl> <list>    
1 character         11  57.9 <chr [11]>
2 logical            3  15.8 <chr [3]> 
3 numeric            3  15.8 <chr [3]> 
4 POSIXct POSIXt     2  10.5 <chr [2]> 
mydata %>% dplyr::select(-keycolumns, -contains("Date")) %>% describer::describe() %>% 
    knitr::kable(format = "markdown")
.column_name .column_class .column_type .count_elements .mean_value .sd_value .q0_value .q25_value .q50_value .q75_value .q100_value
Sex character character 250 NA NA Female NA NA NA Male
Age numeric double 250 49.538153 14.1595015 25 37 49 61 73
Race character character 250 NA NA Asian NA NA NA White
PreinvasiveComponent character character 250 NA NA Absent NA NA NA Present
LVI character character 250 NA NA Absent NA NA NA Present
PNI character character 250 NA NA Absent NA NA NA Present
Death logical logical 250 NA NA FALSE NA NA NA TRUE
Group character character 250 NA NA Control NA NA NA Treatment
Grade character character 250 NA NA 1 NA NA NA 3
TStage character character 250 NA NA 1 NA NA NA 4
AntiX_intensity numeric double 250 2.389558 0.6636071 1 2 2 3 3
AntiY_intensity numeric double 250 2.016064 0.7980211 1 1 2 3 3
LymphNodeMetastasis character character 250 NA NA Absent NA NA NA Present
Valid logical logical 250 NA NA FALSE NA NA NA TRUE
Smoker logical logical 250 NA NA FALSE NA NA NA TRUE
Grade_Level character character 250 NA NA high NA NA NA moderate
DeathTime character character 250 NA NA MoreThan1Year NA NA NA Within1Year

Plot variable types

mydata %>% dplyr::select(-keycolumns) %>% inspectdf::inspect_types() %>% inspectdf::show_plot()

# https://github.com/ropensci/visdat
# http://visdat.njtierney.com/articles/using_visdat.html
# https://cran.r-project.org/web/packages/visdat/index.html
# http://visdat.njtierney.com/

# visdat::vis_guess(mydata)

visdat::vis_dat(mydata)

mydata %>% explore::explore_tbl()

2.7.3 Define Variable Types

2.7.3.1 Find character variables

characterVariables <- mydata %>% dplyr::select(-keycolumns) %>% inspectdf::inspect_types() %>% 
    dplyr::filter(type == "character") %>% dplyr::select(col_name) %>% dplyr::pull() %>% 
    unlist()

characterVariables
 [1] "Sex"                  "Race"                 "PreinvasiveComponent"
 [4] "LVI"                  "PNI"                  "Group"               
 [7] "Grade"                "TStage"               "LymphNodeMetastasis" 
[10] "Grade_Level"          "DeathTime"           

2.7.3.2 Find categorical variables

categoricalVariables <- mydata %>% dplyr::select(-keycolumns, -contains("Date")) %>% 
    describer::describe() %>% janitor::clean_names() %>% dplyr::filter(column_type == 
    "factor") %>% dplyr::select(column_name) %>% dplyr::pull()

categoricalVariables
character(0)

2.7.3.3 Find continious variables

continiousVariables <- mydata %>% dplyr::select(-keycolumns, -contains("Date")) %>% 
    describer::describe() %>% janitor::clean_names() %>% dplyr::filter(column_type == 
    "numeric" | column_type == "double") %>% dplyr::select(column_name) %>% dplyr::pull()

continiousVariables
[1] "Age"             "AntiX_intensity" "AntiY_intensity"

2.7.3.4 Find numeric variables

numericVariables <- mydata %>% dplyr::select(-keycolumns) %>% inspectdf::inspect_types() %>% 
    dplyr::filter(type == "numeric") %>% dplyr::select(col_name) %>% dplyr::pull() %>% 
    unlist()

numericVariables
[1] "Age"             "AntiX_intensity" "AntiY_intensity"

2.7.3.5 Find integer variables

integerVariables <- mydata %>% dplyr::select(-keycolumns) %>% inspectdf::inspect_types() %>% 
    dplyr::filter(type == "integer") %>% dplyr::select(col_name) %>% dplyr::pull() %>% 
    unlist()

integerVariables
NULL

2.7.3.6 Find list variables

listVariables <- mydata %>% dplyr::select(-keycolumns) %>% inspectdf::inspect_types() %>% 
    dplyr::filter(type == "list") %>% dplyr::select(col_name) %>% dplyr::pull() %>% 
    unlist()
listVariables
NULL

2.7.3.7 Find date variables

is_date <- function(x) inherits(x, c("POSIXct", "POSIXt"))

dateVariables <- names(which(sapply(mydata, FUN = is_date) == TRUE))
dateVariables
[1] "LastFollowUpDate" "SurgeryDate"     

2.8 Overview the Data

Codes for overviewing the data.20

2.8.1 View Data

View(mydata)
reactable::reactable(data = mydata, sortable = TRUE, resizable = TRUE, filterable = TRUE, 
    searchable = TRUE, pagination = TRUE, paginationType = "numbers", showPageSizeOptions = TRUE, 
    highlight = TRUE, striped = TRUE, outlined = TRUE, compact = TRUE, wrap = FALSE, 
    showSortIcon = TRUE, showSortable = TRUE)

2.8.2 Overview / Exploratory Data Analysis (EDA)

Summary of Data via summarytools 📦

summarytools::view(summarytools::dfSummary(mydata %>% dplyr::select(-keycolumns)))
if (!dir.exists(here::here("out"))) {
    dir.create(here::here("out"))
}

summarytools::view(x = summarytools::dfSummary(mydata %>% dplyr::select(-keycolumns)), 
    file = here::here("out", "mydata_summary.html"))

Summary via dataMaid 📦

if (!dir.exists(here::here("out"))) {
    dir.create(here::here("out"))
}

dataMaid::makeDataReport(data = mydata, file = here::here("out", "dataMaid_mydata.Rmd"), 
    replace = TRUE, openResult = FALSE, render = FALSE, quiet = TRUE)

Summary via explore 📦

if (!dir.exists(here::here("out"))) {
    dir.create(here::here("out"))
}

mydata %>% dplyr::select(-dateVariables) %>% explore::report(output_file = "mydata_report.html", 
    output_dir = here::here("out"))

Glimpse of Data

dplyr::glimpse(mydata %>% dplyr::select(-keycolumns, -dateVariables))
Observations: 250
Variables: 17
$ Sex                  <chr> "Female", "Female", "Female", "Female", "Male", …
$ Age                  <dbl> 30, 32, 53, 57, 47, 58, 59, 54, 35, 27, 53, 55, …
$ Race                 <chr> "White", "White", "White", "Hispanic", "White", …
$ PreinvasiveComponent <chr> "Absent", "Absent", "Absent", "Absent", "Absent"…
$ LVI                  <chr> "Present", "Absent", "Absent", "Present", "Absen…
$ PNI                  <chr> "Absent", "Absent", "Absent", "Present", "Absent…
$ Death                <lgl> FALSE, TRUE, TRUE, TRUE, TRUE, FALSE, FALSE, TRU…
$ Group                <chr> "Control", "Control", "Control", "Control", "Con…
$ Grade                <chr> "1", "1", "2", "1", "2", "2", "3", "1", "2", "1"…
$ TStage               <chr> "4", "4", "3", "3", "1", "3", "3", "3", "4", "4"…
$ AntiX_intensity      <dbl> 2, 2, 2, 2, 3, 1, 1, 3, 2, 3, 2, 3, 1, 3, 1, 2, …
$ AntiY_intensity      <dbl> 2, 2, 2, 3, 2, 1, 2, 3, 3, 1, 1, 2, 1, 3, 1, 2, …
$ LymphNodeMetastasis  <chr> "Present", "Absent", "Present", "Present", "Pres…
$ Valid                <lgl> TRUE, FALSE, TRUE, TRUE, FALSE, FALSE, TRUE, TRU…
$ Smoker               <lgl> TRUE, TRUE, FALSE, FALSE, FALSE, FALSE, TRUE, TR…
$ Grade_Level          <chr> "moderate", "moderate", "high", "low", "high", "…
$ DeathTime            <chr> "Within1Year", "Within1Year", "Within1Year", "Wi…
mydata %>% explore::describe()
# A tibble: 21 x 8
   variable             type     na na_pct unique   min  mean   max
   <chr>                <chr> <int>  <dbl>  <int> <dbl> <dbl> <dbl>
 1 ID                   chr       0    0      250    NA NA       NA
 2 Name                 chr       1    0.4    250    NA NA       NA
 3 Sex                  chr       1    0.4      3    NA NA       NA
 4 Age                  dbl       1    0.4     50    25 49.5     73
 5 Race                 chr       1    0.4      8    NA NA       NA
 6 PreinvasiveComponent chr       1    0.4      3    NA NA       NA
 7 LVI                  chr       1    0.4      3    NA NA       NA
 8 PNI                  chr       1    0.4      3    NA NA       NA
 9 LastFollowUpDate     dat       1    0.4     13    NA NA       NA
10 Death                lgl       1    0.4      3     0  0.67     1
# … with 11 more rows

Explore

explore::explore(mydata)

2.8.3 Control Data

Control Data if matching expectations

visdat::vis_expect(data = mydata, expectation = ~.x == -1, show_perc = TRUE)

visdat::vis_expect(mydata, ~.x >= 25)

See missing values

visdat::vis_miss(airquality, cluster = TRUE)

visdat::vis_miss(airquality, sort_miss = TRUE)

xray::anomalies(mydata)
$variables
               Variable   q qNA  pNA qZero pZero qBlank pBlank qInf pInf
1                Smoker 250   1 0.4%   130   52%      0      -    0    -
2                 Valid 250   1 0.4%   116 46.4%      0      -    0    -
3                 Death 250   1 0.4%    83 33.2%      0      -    0    -
4                   Sex 250   1 0.4%     0     -      0      -    0    -
5  PreinvasiveComponent 250   1 0.4%     0     -      0      -    0    -
6                   LVI 250   1 0.4%     0     -      0      -    0    -
7                   PNI 250   1 0.4%     0     -      0      -    0    -
8                 Group 250   1 0.4%     0     -      0      -    0    -
9   LymphNodeMetastasis 250   1 0.4%     0     -      0      -    0    -
10                Grade 250   1 0.4%     0     -      0      -    0    -
11      AntiX_intensity 250   1 0.4%     0     -      0      -    0    -
12      AntiY_intensity 250   1 0.4%     0     -      0      -    0    -
13          Grade_Level 250   1 0.4%     0     -      0      -    0    -
14                 Race 250   1 0.4%     0     -      0      -    0    -
15     LastFollowUpDate 250   1 0.4%     0     -      0      -    0    -
16                  Age 250   1 0.4%     0     -      0      -    0    -
17          SurgeryDate 250   1 0.4%     0     -      0      -    0    -
18                 Name 250   1 0.4%     0     -      0      -    0    -
19            DeathTime 250   0    -     0     -      0      -    0    -
20               TStage 250   0    -     0     -      0      -    0    -
21                   ID 250   0    -     0     -      0      -    0    -
   qDistinct      type anomalous_percent
1          3   Logical             52.4%
2          3   Logical             46.8%
3          3   Logical             33.6%
4          3 Character              0.4%
5          3 Character              0.4%
6          3 Character              0.4%
7          3 Character              0.4%
8          3 Character              0.4%
9          3 Character              0.4%
10         4 Character              0.4%
11         4   Numeric              0.4%
12         4   Numeric              0.4%
13         4 Character              0.4%
14         8 Character              0.4%
15        13 Timestamp              0.4%
16        50   Numeric              0.4%
17       233 Timestamp              0.4%
18       250 Character              0.4%
19         2 Character                 -
20         4 Character                 -
21       250 Character                 -

$problem_variables
 [1] Variable          q                 qNA               pNA              
 [5] qZero             pZero             qBlank            pBlank           
 [9] qInf              pInf              qDistinct         type             
[13] anomalous_percent problems         
<0 rows> (or 0-length row.names)
xray::distributions(mydata)
================================================================================

[1] "Ignoring variable LastFollowUpDate: Unsupported type for visualization."

[1] "Ignoring variable SurgeryDate: Unsupported type for visualization."

         Variable p_1 p_10 p_25 p_50 p_75 p_90 p_99
1 AntiX_intensity   1  1.8    2    2    3    3    3
2 AntiY_intensity   1    1    1    2    3    3    3
3             Age  25 30.8   37   49   61   70   73

2.8.4 Explore Data

Summary of Data via DataExplorer 📦

DataExplorer::plot_str(mydata)
DataExplorer::plot_str(mydata, type = "r")
DataExplorer::introduce(mydata)
# A tibble: 1 x 9
   rows columns discrete_columns continuous_colu… all_missing_col…
  <int>   <int>            <int>            <int>            <int>
1   250      21               18                3                0
# … with 4 more variables: total_missing_values <int>, complete_rows <int>,
#   total_observations <int>, memory_usage <dbl>
DataExplorer::plot_intro(mydata)

DataExplorer::plot_missing(mydata)

Drop columns

mydata2 <- DataExplorer::drop_columns(mydata, "TStage")
DataExplorer::plot_bar(mydata)

DataExplorer::plot_bar(mydata, with = "Death")

DataExplorer::plot_histogram(mydata)



3 Statistical Analysis

Learn these tests as highlighted in (Schmidt et al. 2017).21


4 Results

Write results as described in (Knijn, Simmer, and Nagtegaal 2015)22

  • Describe the number of patients included in the analysis and reason for dropout

  • Report patient/disease characteristics (including the biomarker of interest) with the number of missing values

  • Describe the interaction of the biomarker of interest with established prognostic variables

  • Include at least 90 % of initial cases included in univariate and multivariate analyses

  • Report the estimated effect (relative risk/odds ratio, confidence interval, and p value) in univariate analysis

  • Report the estimated effect (hazard rate/odds ratio, confidence interval, and p value) in multivariate analysis

  • Report the estimated effects (hazard ratio/odds ratio, confidence interval, and p value) of other prognostic factors included in multivariate analysis


4.1 Data Dictionary

Codes for generating data dictionary.23


4.2 Clean and Recode Data

Codes for clean and recode data.24

questionr::irec()

questionr::iorder()

questionr::icut()

iris %>% mutate(sumVar = rowSums(.[1:4]))

iris %>% mutate(sumVar = rowSums(select(., contains(“Sepal”)))) %>% head

iris %>% mutate(sumVar = select(., contains(“Sepal”)) %>% rowSums()) %>% head

iRenameColumn.R

iSelectColumn.R

<= 22 Low
>= 23 & <= 41 Average 
>=42 High

4.3 Impute Missing Data

4.4 impute

Codes for missing data and impute.25

4.4.1 Missing Data

Plot missing data

visdat::vis_miss(mydata)

4.4.2 impute continious

4.4.3 impute categorical

4.4.4 impute outlier

4.5 transform

4.5.1 min -max

4.5.2 skewness

4.5.3 log

4.6 binning

4.6.1 optimal binning

4.6.2 standardize

4.7 data transformation report

4.8 inspectdf


4.9 Descriptive Statistics

Codes for Descriptive Statistics.26

4.9.1 Table One

Report Data properties via report 📦

mydata %>% dplyr::select(-dplyr::contains("Date")) %>% report::report()
The data contains 250 observations of the following variables:
  - ID: 250 entries: 001, n = 1; 002, n = 1; 003, n = 1 and 247 others (0 missing)
  - Name: 249 entries: Aceyn, n = 1; Adalaide, n = 1; Adidas, n = 1 and 246 others (1 missing)
  - Sex: 2 entries: Male, n = 127; Female, n = 122 (1 missing)
  - Age: Mean = 49.54, SD = 14.16, Median = , MAD = 17.79, range: [25, 73], Skewness = 0.00, Kurtosis = -1.15, 1 missing
  - Race: 7 entries: White, n = 158; Hispanic, n = 38; Black, n = 30 and 4 others (1 missing)
  - PreinvasiveComponent: 2 entries: Absent, n = 203; Present, n = 46 (1 missing)
  - LVI: 2 entries: Absent, n = 147; Present, n = 102 (1 missing)
  - PNI: 2 entries: Absent, n = 171; Present, n = 78 (1 missing)
  - Death: 2 levels: FALSE (n = 83, 33.20%); TRUE (n = 166, 66.40%) and missing (n = 1, 0.40%)
  - Group: 2 entries: Treatment, n = 131; Control, n = 118 (1 missing)
  - Grade: 3 entries: 3, n = 109; 1, n = 78; 2, n = 62 (1 missing)
  - TStage: 4 entries: 4, n = 118; 3, n = 65; 2, n = 43 and 1 other (0 missing)
  - AntiX_intensity: Mean = 2.39, SD = 0.66, Median = , MAD = 1.48, range: [1, 3], Skewness = -0.63, Kurtosis = -0.65, 1 missing
  - AntiY_intensity: Mean = 2.02, SD = 0.80, Median = , MAD = 1.48, range: [1, 3], Skewness = -0.03, Kurtosis = -1.42, 1 missing
  - LymphNodeMetastasis: 2 entries: Absent, n = 144; Present, n = 105 (1 missing)
  - Valid: 2 levels: FALSE (n = 116, 46.40%); TRUE (n = 133, 53.20%) and missing (n = 1, 0.40%)
  - Smoker: 2 levels: FALSE (n = 130, 52.00%); TRUE (n = 119, 47.60%) and missing (n = 1, 0.40%)
  - Grade_Level: 3 entries: high, n = 109; low, n = 77; moderate, n = 63 (1 missing)
  - DeathTime: 2 entries: Within1Year, n = 149; MoreThan1Year, n = 101 (0 missing)

Table 1 via arsenal 📦

# cat(names(mydata), sep = " + \n")
library(arsenal)
tab1 <- arsenal::tableby(
  ~ Sex +
    Age +
    Race +
    PreinvasiveComponent +
    LVI +
    PNI +
    Death +
    Group +
    Grade +
    TStage +
    # `Anti-X-intensity` +
    # `Anti-Y-intensity` +
    LymphNodeMetastasis +
    Valid +
    Smoker +
    Grade_Level
  ,
  data = mydata 
)
summary(tab1)
Overall (N=250)
Sex
   N-Miss 1
   Female 122 (49.0%)
   Male 127 (51.0%)
Age
   N-Miss 1
   Mean (SD) 49.538 (14.160)
   Range 25.000 - 73.000
Race
   N-Miss 1
   Asian 15 (6.0%)
   Bi-Racial 5 (2.0%)
   Black 30 (12.0%)
   Hispanic 38 (15.3%)
   Native 2 (0.8%)
   Other 1 (0.4%)
   White 158 (63.5%)
PreinvasiveComponent
   N-Miss 1
   Absent 203 (81.5%)
   Present 46 (18.5%)
LVI
   N-Miss 1
   Absent 147 (59.0%)
   Present 102 (41.0%)
PNI
   N-Miss 1
   Absent 171 (68.7%)
   Present 78 (31.3%)
Death
   N-Miss 1
   FALSE 83 (33.3%)
   TRUE 166 (66.7%)
Group
   N-Miss 1
   Control 118 (47.4%)
   Treatment 131 (52.6%)
Grade
   N-Miss 1
   1 78 (31.3%)
   2 62 (24.9%)
   3 109 (43.8%)
TStage
   1 24 (9.6%)
   2 43 (17.2%)
   3 65 (26.0%)
   4 118 (47.2%)
LymphNodeMetastasis
   N-Miss 1
   Absent 144 (57.8%)
   Present 105 (42.2%)
Valid
   N-Miss 1
   FALSE 116 (46.6%)
   TRUE 133 (53.4%)
Smoker
   N-Miss 1
   FALSE 130 (52.2%)
   TRUE 119 (47.8%)
Grade_Level
   N-Miss 1
   high 109 (43.8%)
   low 77 (30.9%)
   moderate 63 (25.3%)

Table 1 via tableone 📦

library(tableone)
mydata %>% dplyr::select(-keycolumns, -dateVariables) %>% tableone::CreateTableOne(data = .)
                                    
                                     Overall      
  n                                    250        
  Sex = Male (%)                       127 (51.0) 
  Age (mean (SD))                    49.54 (14.16)
  Race (%)                                        
     Asian                              15 ( 6.0) 
     Bi-Racial                           5 ( 2.0) 
     Black                              30 (12.0) 
     Hispanic                           38 (15.3) 
     Native                              2 ( 0.8) 
     Other                               1 ( 0.4) 
     White                             158 (63.5) 
  PreinvasiveComponent = Present (%)    46 (18.5) 
  LVI = Present (%)                    102 (41.0) 
  PNI = Present (%)                     78 (31.3) 
  Death = TRUE (%)                     166 (66.7) 
  Group = Treatment (%)                131 (52.6) 
  Grade (%)                                       
     1                                  78 (31.3) 
     2                                  62 (24.9) 
     3                                 109 (43.8) 
  TStage (%)                                      
     1                                  24 ( 9.6) 
     2                                  43 (17.2) 
     3                                  65 (26.0) 
     4                                 118 (47.2) 
  AntiX_intensity (mean (SD))         2.39 (0.66) 
  AntiY_intensity (mean (SD))         2.02 (0.80) 
  LymphNodeMetastasis = Present (%)    105 (42.2) 
  Valid = TRUE (%)                     133 (53.4) 
  Smoker = TRUE (%)                    119 (47.8) 
  Grade_Level (%)                                 
     high                              109 (43.8) 
     low                                77 (30.9) 
     moderate                           63 (25.3) 
  DeathTime = Within1Year (%)          149 (59.6) 

Descriptive Statistics of Continuous Variables

mydata %>% dplyr::select(continiousVariables, numericVariables, integerVariables) %>% 
    summarytools::descr(., style = "rmarkdown")
print(summarytools::descr(mydata), method = "render", table.classes = "st-small")
mydata %>% summarytools::descr(., stats = "common", transpose = TRUE, headings = FALSE)
mydata %>% summarytools::descr(stats = "common") %>% summarytools::tb()
mydata$Sex %>% summarytools::freq(cumul = FALSE, report.nas = FALSE) %>% summarytools::tb()
mydata %>% explore::describe() %>% dplyr::filter(unique < 5)
# A tibble: 15 x 8
   variable             type     na na_pct unique   min  mean   max
   <chr>                <chr> <int>  <dbl>  <int> <dbl> <dbl> <dbl>
 1 Sex                  chr       1    0.4      3    NA NA       NA
 2 PreinvasiveComponent chr       1    0.4      3    NA NA       NA
 3 LVI                  chr       1    0.4      3    NA NA       NA
 4 PNI                  chr       1    0.4      3    NA NA       NA
 5 Death                lgl       1    0.4      3     0  0.67     1
 6 Group                chr       1    0.4      3    NA NA       NA
 7 Grade                chr       1    0.4      4    NA NA       NA
 8 TStage               chr       0    0        4    NA NA       NA
 9 AntiX_intensity      dbl       1    0.4      4     1  2.39     3
10 AntiY_intensity      dbl       1    0.4      4     1  2.02     3
11 LymphNodeMetastasis  chr       1    0.4      3    NA NA       NA
12 Valid                lgl       1    0.4      3     0  0.53     1
13 Smoker               lgl       1    0.4      3     0  0.48     1
14 Grade_Level          chr       1    0.4      4    NA NA       NA
15 DeathTime            chr       0    0        2    NA NA       NA
mydata %>% explore::describe() %>% dplyr::filter(na > 0)
# A tibble: 18 x 8
   variable             type     na na_pct unique   min  mean   max
   <chr>                <chr> <int>  <dbl>  <int> <dbl> <dbl> <dbl>
 1 Name                 chr       1    0.4    250    NA NA       NA
 2 Sex                  chr       1    0.4      3    NA NA       NA
 3 Age                  dbl       1    0.4     50    25 49.5     73
 4 Race                 chr       1    0.4      8    NA NA       NA
 5 PreinvasiveComponent chr       1    0.4      3    NA NA       NA
 6 LVI                  chr       1    0.4      3    NA NA       NA
 7 PNI                  chr       1    0.4      3    NA NA       NA
 8 LastFollowUpDate     dat       1    0.4     13    NA NA       NA
 9 Death                lgl       1    0.4      3     0  0.67     1
10 Group                chr       1    0.4      3    NA NA       NA
11 Grade                chr       1    0.4      4    NA NA       NA
12 AntiX_intensity      dbl       1    0.4      4     1  2.39     3
13 AntiY_intensity      dbl       1    0.4      4     1  2.02     3
14 LymphNodeMetastasis  chr       1    0.4      3    NA NA       NA
15 Valid                lgl       1    0.4      3     0  0.53     1
16 Smoker               lgl       1    0.4      3     0  0.48     1
17 Grade_Level          chr       1    0.4      4    NA NA       NA
18 SurgeryDate          dat       1    0.4    233    NA NA       NA
mydata %>% explore::describe()
# A tibble: 21 x 8
   variable             type     na na_pct unique   min  mean   max
   <chr>                <chr> <int>  <dbl>  <int> <dbl> <dbl> <dbl>
 1 ID                   chr       0    0      250    NA NA       NA
 2 Name                 chr       1    0.4    250    NA NA       NA
 3 Sex                  chr       1    0.4      3    NA NA       NA
 4 Age                  dbl       1    0.4     50    25 49.5     73
 5 Race                 chr       1    0.4      8    NA NA       NA
 6 PreinvasiveComponent chr       1    0.4      3    NA NA       NA
 7 LVI                  chr       1    0.4      3    NA NA       NA
 8 PNI                  chr       1    0.4      3    NA NA       NA
 9 LastFollowUpDate     dat       1    0.4     13    NA NA       NA
10 Death                lgl       1    0.4      3     0  0.67     1
# … with 11 more rows

4.9.2 Categorical Variables

Use R/gc_desc_cat.R to generate gc_desc_cat.Rmd containing descriptive statistics for categorical variables

source(here::here("R", "gc_desc_cat.R"))

4.9.2.1 Descriptive Statistics Sex

mydata %>% janitor::tabyl(Sex) %>% janitor::adorn_pct_formatting(rounding = "half up", 
    digits = 1) %>% knitr::kable()
Sex n percent valid_percent
Female 122 48.8% 49.0%
Male 127 50.8% 51.0%
NA 1 0.4% -

4.9.2.2 Descriptive Statistics Race

mydata %>% janitor::tabyl(Race) %>% janitor::adorn_pct_formatting(rounding = "half up", 
    digits = 1) %>% knitr::kable()
Race n percent valid_percent
Asian 15 6.0% 6.0%
Bi-Racial 5 2.0% 2.0%
Black 30 12.0% 12.0%
Hispanic 38 15.2% 15.3%
Native 2 0.8% 0.8%
Other 1 0.4% 0.4%
White 158 63.2% 63.5%
NA 1 0.4% -

4.9.2.3 Descriptive Statistics PreinvasiveComponent

mydata %>% janitor::tabyl(PreinvasiveComponent) %>% janitor::adorn_pct_formatting(rounding = "half up", 
    digits = 1) %>% knitr::kable()
PreinvasiveComponent n percent valid_percent
Absent 203 81.2% 81.5%
Present 46 18.4% 18.5%
NA 1 0.4% -

4.9.2.4 Descriptive Statistics LVI

mydata %>% janitor::tabyl(LVI) %>% janitor::adorn_pct_formatting(rounding = "half up", 
    digits = 1) %>% knitr::kable()
LVI n percent valid_percent
Absent 147 58.8% 59.0%
Present 102 40.8% 41.0%
NA 1 0.4% -

4.9.2.5 Descriptive Statistics PNI

mydata %>% janitor::tabyl(PNI) %>% janitor::adorn_pct_formatting(rounding = "half up", 
    digits = 1) %>% knitr::kable()
PNI n percent valid_percent
Absent 171 68.4% 68.7%
Present 78 31.2% 31.3%
NA 1 0.4% -

4.9.2.6 Descriptive Statistics Group

mydata %>% janitor::tabyl(Group) %>% janitor::adorn_pct_formatting(rounding = "half up", 
    digits = 1) %>% knitr::kable()
Group n percent valid_percent
Control 118 47.2% 47.4%
Treatment 131 52.4% 52.6%
NA 1 0.4% -

4.9.2.7 Descriptive Statistics Grade

mydata %>% janitor::tabyl(Grade) %>% janitor::adorn_pct_formatting(rounding = "half up", 
    digits = 1) %>% knitr::kable()
Grade n percent valid_percent
1 78 31.2% 31.3%
2 62 24.8% 24.9%
3 109 43.6% 43.8%
NA 1 0.4% -

4.9.2.8 Descriptive Statistics TStage

mydata %>% janitor::tabyl(TStage) %>% janitor::adorn_pct_formatting(rounding = "half up", 
    digits = 1) %>% knitr::kable()
TStage n percent
1 24 9.6%
2 43 17.2%
3 65 26.0%
4 118 47.2%

4.9.2.9 Descriptive Statistics LymphNodeMetastasis

mydata %>% janitor::tabyl(LymphNodeMetastasis) %>% janitor::adorn_pct_formatting(rounding = "half up", 
    digits = 1) %>% knitr::kable()
LymphNodeMetastasis n percent valid_percent
Absent 144 57.6% 57.8%
Present 105 42.0% 42.2%
NA 1 0.4% -

4.9.2.10 Descriptive Statistics Grade_Level

mydata %>% janitor::tabyl(Grade_Level) %>% janitor::adorn_pct_formatting(rounding = "half up", 
    digits = 1) %>% knitr::kable()
Grade_Level n percent valid_percent
high 109 43.6% 43.8%
low 77 30.8% 30.9%
moderate 63 25.2% 25.3%
NA 1 0.4% -

4.9.2.11 Descriptive Statistics DeathTime

mydata %>% janitor::tabyl(DeathTime) %>% janitor::adorn_pct_formatting(rounding = "half up", 
    digits = 1) %>% knitr::kable()
DeathTime n percent
MoreThan1Year 101 40.4%
Within1Year 149 59.6%
race_stats <- summarytools::freq(mydata$Race)
print(race_stats, report.nas = FALSE, totals = FALSE, display.type = FALSE, Variable.label = "Race Group")
mydata %>% explore::describe(PreinvasiveComponent)
variable = PreinvasiveComponent
type     = character
na       = 1 of 250 (0.4%)
unique   = 3
 Absent  = 203 (81.2%)
 Present = 46 (18.4%)
 NA      = 1 (0.4%)
## Frequency or custom tables for categorical variables
SmartEDA::ExpCTable(mydata, Target = NULL, margin = 1, clim = 10, nlim = 5, round = 2, 
    bin = NULL, per = T)
               Variable         Valid Frequency Percent CumPercent
1                   Sex        Female       122    48.8       48.8
2                   Sex          Male       127    50.8       99.6
3                   Sex            NA         1     0.4      100.0
4                   Sex         TOTAL       250      NA         NA
5                  Race         Asian        15     6.0        6.0
6                  Race     Bi-Racial         5     2.0        8.0
7                  Race         Black        30    12.0       20.0
8                  Race      Hispanic        38    15.2       35.2
9                  Race            NA         1     0.4       35.6
10                 Race        Native         2     0.8       36.4
11                 Race         Other         1     0.4       36.8
12                 Race         White       158    63.2      100.0
13                 Race         TOTAL       250      NA         NA
14 PreinvasiveComponent        Absent       203    81.2       81.2
15 PreinvasiveComponent            NA         1     0.4       81.6
16 PreinvasiveComponent       Present        46    18.4      100.0
17 PreinvasiveComponent         TOTAL       250      NA         NA
18                  LVI        Absent       147    58.8       58.8
19                  LVI            NA         1     0.4       59.2
20                  LVI       Present       102    40.8      100.0
21                  LVI         TOTAL       250      NA         NA
22                  PNI        Absent       171    68.4       68.4
23                  PNI            NA         1     0.4       68.8
24                  PNI       Present        78    31.2      100.0
25                  PNI         TOTAL       250      NA         NA
26                Group       Control       118    47.2       47.2
27                Group            NA         1     0.4       47.6
28                Group     Treatment       131    52.4      100.0
29                Group         TOTAL       250      NA         NA
30                Grade             1        78    31.2       31.2
31                Grade             2        62    24.8       56.0
32                Grade             3       109    43.6       99.6
33                Grade            NA         1     0.4      100.0
34                Grade         TOTAL       250      NA         NA
35               TStage             1        24     9.6        9.6
36               TStage             2        43    17.2       26.8
37               TStage             3        65    26.0       52.8
38               TStage             4       118    47.2      100.0
39               TStage         TOTAL       250      NA         NA
40  LymphNodeMetastasis        Absent       144    57.6       57.6
41  LymphNodeMetastasis            NA         1     0.4       58.0
42  LymphNodeMetastasis       Present       105    42.0      100.0
43  LymphNodeMetastasis         TOTAL       250      NA         NA
44          Grade_Level          high       109    43.6       43.6
45          Grade_Level           low        77    30.8       74.4
46          Grade_Level      moderate        63    25.2       99.6
47          Grade_Level            NA         1     0.4      100.0
48          Grade_Level         TOTAL       250      NA         NA
49            DeathTime MoreThan1Year       101    40.4       40.4
50            DeathTime   Within1Year       149    59.6      100.0
51            DeathTime         TOTAL       250      NA         NA
52      AntiX_intensity             1        25    10.0       10.0
53      AntiX_intensity             2       102    40.8       50.8
54      AntiX_intensity             3       122    48.8       99.6
55      AntiX_intensity            NA         1     0.4      100.0
56      AntiX_intensity         TOTAL       250      NA         NA
57      AntiY_intensity             1        77    30.8       30.8
58      AntiY_intensity             2        91    36.4       67.2
59      AntiY_intensity             3        81    32.4       99.6
60      AntiY_intensity            NA         1     0.4      100.0
61      AntiY_intensity         TOTAL       250      NA         NA
inspectdf::inspect_cat(mydata)
# A tibble: 16 x 5
   col_name               cnt common      common_pcnt levels            
   <chr>                <int> <chr>             <dbl> <named list>      
 1 Death                    3 TRUE               66.4 <tibble [3 × 3]>  
 2 DeathTime                2 Within1Year        59.6 <tibble [2 × 3]>  
 3 Grade                    4 3                  43.6 <tibble [4 × 3]>  
 4 Grade_Level              4 high               43.6 <tibble [4 × 3]>  
 5 Group                    3 Treatment          52.4 <tibble [3 × 3]>  
 6 ID                     250 001                 0.4 <tibble [250 × 3]>
 7 LVI                      3 Absent             58.8 <tibble [3 × 3]>  
 8 LymphNodeMetastasis      3 Absent             57.6 <tibble [3 × 3]>  
 9 Name                   250 Aceyn               0.4 <tibble [250 × 3]>
10 PNI                      3 Absent             68.4 <tibble [3 × 3]>  
11 PreinvasiveComponent     3 Absent             81.2 <tibble [3 × 3]>  
12 Race                     8 White              63.2 <tibble [8 × 3]>  
13 Sex                      3 Male               50.8 <tibble [3 × 3]>  
14 Smoker                   3 FALSE              52   <tibble [3 × 3]>  
15 TStage                   4 4                  47.2 <tibble [4 × 3]>  
16 Valid                    3 TRUE               53.2 <tibble [3 × 3]>  
inspectdf::inspect_cat(mydata)$levels$Group
# A tibble: 3 x 3
  value      prop   cnt
  <chr>     <dbl> <int>
1 Treatment 0.524   131
2 Control   0.472   118
3 <NA>      0.004     1

4.9.2.12 Split-Group Stats Categorical

library(summarytools)

grouped_freqs <- stby(data = mydata$Smoker, INDICES = mydata$Sex, FUN = freq, cumul = FALSE, 
    report.nas = FALSE)

grouped_freqs %>% tb(order = 2)

4.9.2.13 Grouped Categorical

summarytools::stby(list(x = mydata$LVI, y = mydata$LymphNodeMetastasis), mydata$PNI, 
    summarytools::ctable)
with(mydata, summarytools::stby(list(x = LVI, y = LymphNodeMetastasis), PNI, summarytools::ctable))
mydata %>% dplyr::select(characterVariables) %>% dplyr::select(PreinvasiveComponent, 
    PNI, LVI) %>% reactable::reactable(data = ., groupBy = c("PreinvasiveComponent", 
    "PNI"), columns = list(LVI = reactable::colDef(aggregate = "count")))

4.9.3 Continious Variables

questionr:::icut()
source(here::here("R", "gc_desc_cont.R"))

Descriptive Statistics Age

mydata %>% jmv::descriptives(data = ., vars = "Age", hist = TRUE, dens = TRUE, box = TRUE, 
    violin = TRUE, dot = TRUE, mode = TRUE, sd = TRUE, variance = TRUE, skew = TRUE, 
    kurt = TRUE, quart = TRUE)

 DESCRIPTIVES

 Descriptives                       
 ────────────────────────────────── 
                          Age       
 ────────────────────────────────── 
   N                          249   
   Missing                      1   
   Mean                      49.5   
   Median                    49.0   
   Mode                      72.0   
   Standard deviation        14.2   
   Variance                   200   
   Minimum                   25.0   
   Maximum                   73.0   
   Skewness               0.00389   
   Std. error skewness      0.154   
   Kurtosis                 -1.15   
   Std. error kurtosis      0.307   
   25th percentile           37.0   
   50th percentile           49.0   
   75th percentile           61.0   
 ────────────────────────────────── 

Descriptive Statistics AntiX_intensity

mydata %>% jmv::descriptives(data = ., vars = "AntiX_intensity", hist = TRUE, dens = TRUE, 
    box = TRUE, violin = TRUE, dot = TRUE, mode = TRUE, sd = TRUE, variance = TRUE, 
    skew = TRUE, kurt = TRUE, quart = TRUE)

 DESCRIPTIVES

 Descriptives                               
 ────────────────────────────────────────── 
                          AntiX_intensity   
 ────────────────────────────────────────── 
   N                                  249   
   Missing                              1   
   Mean                              2.39   
   Median                            2.00   
   Mode                              3.00   
   Standard deviation               0.664   
   Variance                         0.440   
   Minimum                           1.00   
   Maximum                           3.00   
   Skewness                        -0.631   
   Std. error skewness              0.154   
   Kurtosis                        -0.640   
   Std. error kurtosis              0.307   
   25th percentile                   2.00   
   50th percentile                   2.00   
   75th percentile                   3.00   
 ────────────────────────────────────────── 

Descriptive Statistics AntiY_intensity

mydata %>% jmv::descriptives(data = ., vars = "AntiY_intensity", hist = TRUE, dens = TRUE, 
    box = TRUE, violin = TRUE, dot = TRUE, mode = TRUE, sd = TRUE, variance = TRUE, 
    skew = TRUE, kurt = TRUE, quart = TRUE)

 DESCRIPTIVES

 Descriptives                               
 ────────────────────────────────────────── 
                          AntiY_intensity   
 ────────────────────────────────────────── 
   N                                  249   
   Missing                              1   
   Mean                              2.02   
   Median                            2.00   
   Mode                              2.00   
   Standard deviation               0.798   
   Variance                         0.637   
   Minimum                           1.00   
   Maximum                           3.00   
   Skewness                       -0.0289   
   Std. error skewness              0.154   
   Kurtosis                         -1.43   
   Std. error kurtosis              0.307   
   25th percentile                   1.00   
   50th percentile                   2.00   
   75th percentile                   3.00   
 ────────────────────────────────────────── 

tab <- tableone::CreateTableOne(data = mydata)
# ?print.ContTable
tab$ContTable
                             
                              Overall      
  n                           250          
  Age (mean (SD))             49.54 (14.16)
  AntiX_intensity (mean (SD))  2.39 (0.66) 
  AntiY_intensity (mean (SD))  2.02 (0.80) 
print(tab$ContTable, nonnormal = c("Anti-X-intensity"))
                             
                              Overall      
  n                           250          
  Age (mean (SD))             49.54 (14.16)
  AntiX_intensity (mean (SD))  2.39 (0.66) 
  AntiY_intensity (mean (SD))  2.02 (0.80) 
mydata %>% explore::describe(Age)
variable = Age
type     = double
na       = 1 of 250 (0.4%)
unique   = 50
min|max  = 25 | 73
q05|q95  = 28 | 72
q25|q75  = 37 | 61
median   = 49
mean     = 49.53815
mydata %>% dplyr::select(continiousVariables) %>% SmartEDA::ExpNumStat(data = ., 
    by = "A", gp = NULL, Qnt = seq(0, 1, 0.1), MesofShape = 2, Outlier = TRUE, round = 2)
inspectdf::inspect_num(mydata, breaks = 10)
# A tibble: 3 x 10
  col_name        min    q1 median  mean    q3   max     sd pcnt_na hist        
  <chr>         <dbl> <dbl>  <dbl> <dbl> <dbl> <dbl>  <dbl>   <dbl> <named list>
1 Age              25    37     49 49.5     61    73 14.2       0.4 <tibble [12…
2 AntiX_intens…     1     2      2  2.39     3     3  0.664     0.4 <tibble [12…
3 AntiY_intens…     1     1      2  2.02     3     3  0.798     0.4 <tibble [12…
inspectdf::inspect_num(mydata)$hist$Age
# A tibble: 27 x 2
   value        prop
   <chr>       <dbl>
 1 [-Inf, 24) 0     
 2 [24, 26)   0.0201
 3 [26, 28)   0.0281
 4 [28, 30)   0.0361
 5 [30, 32)   0.0361
 6 [32, 34)   0.0602
 7 [34, 36)   0.0482
 8 [36, 38)   0.0241
 9 [38, 40)   0.0161
10 [40, 42)   0.0602
# … with 17 more rows
inspectdf::inspect_num(mydata, breaks = 10) %>% inspectdf::show_plot()

4.9.3.1 Split-Group Stats Continious

grouped_descr <- summarytools::stby(data = mydata, INDICES = mydata$Sex, FUN = summarytools::descr, 
    stats = "common")
# grouped_descr %>% summarytools::tb(order = 2)
grouped_descr %>% summarytools::tb()

4.9.3.2 Grouped Continious

summarytools::stby(data = mydata, INDICES = mydata$PreinvasiveComponent, FUN = summarytools::descr, 
    stats = c("mean", "sd", "min", "med", "max"), transpose = TRUE)
with(mydata, summarytools::stby(Age, PreinvasiveComponent, summarytools::descr), 
    stats = c("mean", "sd", "min", "med", "max"), transpose = TRUE)
mydata %>% group_by(PreinvasiveComponent) %>% summarytools::descr(stats = "fivenum")
## Summary statistics by – category
SmartEDA::ExpNumStat(mydata, by = "GA", gp = "PreinvasiveComponent", Qnt = seq(0, 
    1, 0.1), MesofShape = 2, Outlier = TRUE, round = 2)
  Vname                        Group  TN nNeg nZero nPos NegInf PosInf NA_Value
1   Age     PreinvasiveComponent:All 250    0     0  249      0      0        1
2   Age  PreinvasiveComponent:Absent 203    0     0  203      0      0        0
3   Age PreinvasiveComponent:Present  46    0     0   45      0      0        1
4   Age      PreinvasiveComponent:NA   0    0     0    0      0      0        0
  Per_of_Missing   sum min  max  mean median    SD   CV  IQR Skewness Kurtosis
1           0.40 12335  25   73 49.54     49 14.16 0.29 24.0     0.00    -1.16
2           0.00 10117  25   73 49.84     51 14.34 0.29 23.5    -0.02    -1.20
3           2.17  2170  25   72 48.22     49 13.55 0.28 22.0     0.08    -0.98
4            NaN     0 Inf -Inf   NaN     NA    NA   NA   NA      NaN      NaN
  0%  10%  20%  30%  40% 50%  60%  70% 80%  90% 100% LB.25% UB.75% nOutliers
1 25 30.8 34.0 40.4 45.0  49 54.0 59.0  64 70.0   73   1.00  97.00         0
2 25 31.0 34.0 40.6 45.0  51 54.0 59.0  65 70.8   73   2.25  96.25         0
3 25 30.8 34.8 40.2 43.6  49 51.8 56.8  59 68.6   72   3.00  91.00         0
4 NA   NA   NA   NA   NA  NA   NA   NA  NA   NA   NA     NA     NA         0

4.10 Cross Tables

Codes for cross tables.27

library(finalfit)
# dependent <- c('dependent1', 'dependent2' )

# explanatory <- c('explanatory1', 'explanatory2' )

dependent <- "PreinvasiveComponent"

explanatory <- c("Sex", "Age", "Grade", "TStage")

Change column = TRUE argument to get row or column percentages.

source(here::here("R", "gc_table_cross.R"))

Cross Table PreinvasiveComponent

mydata %>%
    summary_factorlist(dependent = 'PreinvasiveComponent', 
                       explanatory = explanatory,
                       # column = TRUE,
                       total_col = TRUE,
                       p = TRUE,
                       add_dependent_label = TRUE,
                       na_include=FALSE
                       # catTest = catTestfisher
                       ) -> table

knitr::kable(table, row.names = FALSE, align = c('l', 'l', 'r', 'r', 'r'))
Dependent: PreinvasiveComponent Absent Present Total p
Sex Female 104 (51.2) 17 (37.8) 121 (48.8) 0.102
Male 99 (48.8) 28 (62.2) 127 (51.2)
Age Mean (SD) 49.8 (14.3) 48.2 (13.6) 49.5 (14.2) 0.492
Grade 1 68 (33.7) 9 (19.6) 77 (31.0) 0.100
2 46 (22.8) 16 (34.8) 62 (25.0)
3 88 (43.6) 21 (45.7) 109 (44.0)
TStage 1 18 (8.9) 6 (13.0) 24 (9.6) 0.117
2 38 (18.7) 4 (8.7) 42 (16.9)
3 48 (23.6) 17 (37.0) 65 (26.1)
4 99 (48.8) 19 (41.3) 118 (47.4)
library(DT)
datatable(mtcars, rownames = FALSE, filter="top", options = list(pageLength = 5, scrollX=T) )

4.10.1 chi-square posthoc pairwise

4.10.1.1 rmngb

4.10.1.2 RVAideMemoire

4.11 Plots

Codes for generating Plots.28

4.12 Categorical Variables

4.12.1 Plots

4.13 Continious Variables

4.13.1 Interactive graphics


R allows to build any type of interactive graphic. My favourite library is plotly that will turn any of your ggplot2 graphic interactive in one supplementary line of code. Try to hover points, to select a zone, to click on the legend.

library(ggplot2)
library(plotly)
library(gapminder)

p <- gapminder %>% filter(year == 1977) %>% ggplot(aes(gdpPercap, lifeExp, size = pop, 
    color = continent)) + geom_point() + scale_x_log10() + theme_bw()

ggplotly(p)

scales::show_col(colours(), cex_label = 0.35)

embedgist <- gistr::gist("https://gist.github.com/sbalci/834ebc154c0ffcb7d5899c42dd3ab75e") %>% 
    gistr::embed()

4.13.2 Alluvial

# https://stackoverflow.com/questions/43053375/weighted-sankey-alluvial-diagram-for-visualizing-discrete-and-continuous-panel/48133004

library(tidyr)
library(dplyr)
library(alluvial)
library(ggplot2)
library(forcats)

set.seed(42)
individual <- rep(LETTERS[1:10], each = 2)
timeperiod <- paste0("time_", rep(1:2, 10))
discretechoice <- factor(paste0("choice_", sample(letters[1:3], 20, replace = T)))
continuouschoice <- ceiling(runif(20, 0, 100))
d <- data.frame(individual, timeperiod, discretechoice, continuouschoice)
# stacked bar diagram of discrete choice by individual
g <- ggplot(data = d, aes(timeperiod, fill = fct_rev(discretechoice)))
g + geom_bar(position = "stack") + guides(fill = guide_legend(title = NULL))

# alluvial diagram of discrete choice by individual
d_alluvial <- d %>% select(individual, timeperiod, discretechoice) %>% spread(timeperiod, 
    discretechoice) %>% group_by(time_1, time_2) %>% summarize(count = n()) %>% ungroup()
Error in UseMethod("ungroup"): no applicable method for 'ungroup' applied to an object of class "list"
alluvial(select(d_alluvial, -count), freq = d_alluvial$count)
Error in log_select(.data, .fun = dplyr::select, .funname = "select", : object 'd_alluvial' not found
# stacked bar diagram of discrete choice, weighting by continuous choice
g + geom_bar(position = "stack", aes(weight = continuouschoice))

library(ggalluvial)
ggplot(data = d, aes(x = timeperiod, stratum = discretechoice, alluvium = individual, 
    y = continuouschoice)) + geom_stratum(aes(fill = discretechoice)) + geom_flow()

CD44changes <- mydata %>% dplyr::select(TumorCD44, TomurcukCD44, PeritumoralTomurcukGr4) %>% 
    dplyr::filter(complete.cases(.)) %>% dplyr::group_by(TumorCD44, TomurcukCD44, 
    PeritumoralTomurcukGr4) %>% dplyr::tally()
Error: Can't subset columns that don't exist.
x The column `TumorCD44` doesn't exist.
library(ggalluvial)

ggplot(data = CD44changes, aes(axis1 = TumorCD44, axis2 = TomurcukCD44, y = n)) + 
    scale_x_discrete(limits = c("TumorCD44", "TomurcukCD44"), expand = c(0.1, 0.05)) + 
    xlab("Tumor Tomurcuk") + geom_alluvium(aes(fill = PeritumoralTomurcukGr4, colour = PeritumoralTomurcukGr4)) + 
    geom_stratum(alpha = 0.5) + geom_text(stat = "stratum", infer.label = TRUE) + 
    # geom_text(stat = 'alluvium', infer.label = TRUE) +
theme_minimal() + ggtitle("Changes in CD44")
Error in ggplot(data = CD44changes, aes(axis1 = TumorCD44, axis2 = TomurcukCD44, : object 'CD44changes' not found

Codes for generating paired tests.29

Codes for generating hypothesis tests.30

5 Hypothesis Tests

5.1 Tests of Normality

5.2 jamovi

mytable <- jmv::ttestIS(formula = HindexCTLA4 ~ PeritumoralTomurcukGr4, data = mydata, 
    vars = HindexCTLA4, students = FALSE, mann = TRUE, norm = TRUE, meanDiff = TRUE, 
    desc = TRUE, plots = TRUE)
Error: Argument 'vars' contains 'HindexCTLA4' which is not present in the dataset
cat("<pre class='jamovitable'>")

```r
print(jtable(mytable$ttest))
```

```
Error in lapply(X = X, FUN = FUN, ...): object 'mytable' not found
```

```r
cat("

")


</pre>




## Categorical







### Chi-Square Cramer Association Predictive Power












## Continious




https://stat.ethz.ch/R-manual/R-devel/library/stats/html/t.test.html

t.test(mtcars$mpg ~ mtcars$am) %>% 
  report::report()


report(t.test(iris$Sepal.Length, iris$Petal.Length))



## Odds








# Frequently Used Statistical Tests By Pathologists

Frequently Used Statistical Tests^[Statistical Literacy Among Academic Pathologists: A Survey Study to Gauge Knowledge of Frequently Used Statistical Tests Among Trainees and Faculty. Archives of Pathology & Laboratory Medicine: February 2017, Vol. 141, No. 2, pp. 279-287. https://doi.org/10.5858/arpa.2016-0200-OA] by [@Schmidt2017]

- Student t test
- Regression/ANOVA
- Chi-square test
- Mann-Whitney test (rank sum)
- Fisher exact test
- Survival analysis
    - Kaplan-Meier/log-rank
    - Cox regression
- Multiple comparison adjustment
    - Tukey
    - Bonferroni
    - Newman-Keuls
- Kappa Statistic
- ROC analysis
- Logistic regression
- Spearman rank correlation
- Kruskal-Wallis test
- Pearson correlation statistic
- Normality test
- McNemar test

# Consider Adding:
- https://cran.r-project.org/web/packages/sm/sm.pdf
- https://cran.r-project.org/web/packages/Rfit/Rfit.pdf


\newpage
\blandscape




**Codes for ROC**.^[See [`childRmd/_16ROC.Rmd`](https://github.com/sbalci/histopathology-template/blob/master/childRmd/_16ROC.Rmd) file for other codes]


# ROC











































**Codes for Decision Tree**.^[See [`childRmd/_17decisionTree.Rmd`](https://github.com/sbalci/histopathology-template/blob/master/childRmd/_17decisionTree.Rmd)]


# Decision Tree







**Explore**


```r
explore::explore(mydata)

5.3 Survival Analysis

Codes for Survival Analysis31

  • Survival analysis with strata, clusters, frailties and competing risks in in Finalfit

https://www.datasurg.net/2019/09/12/survival-analysis-with-strata-clusters-frailties-and-competing-risks-in-in-finalfit/

  • Intracranial WHO grade I meningioma: a competing risk analysis of progression and disease-specific survival

https://link.springer.com/article/10.1007/s00701-019-04096-9

Calculate survival time

mydata$int <- lubridate::interval(lubridate::ymd(mydata$SurgeryDate), lubridate::ymd(mydata$LastFollowUpDate))
mydata$OverallTime <- lubridate::time_length(mydata$int, "month")
mydata$OverallTime <- round(mydata$OverallTime, digits = 1)

recode death status outcome as numbers for survival analysis

## Recoding mydata$Death into mydata$Outcome
mydata$Outcome <- forcats::fct_recode(as.character(mydata$Death), `1` = "TRUE", `0` = "FALSE")
mydata$Outcome <- as.numeric(as.character(mydata$Outcome))

it is always a good practice to double-check after recoding32

table(mydata$Death, mydata$Outcome)
       
          0   1
  FALSE  83   0
  TRUE    0 166

5.3.1 Kaplan-Meier

library(survival)
# data(lung) km <- with(lung, Surv(time, status))
km <- with(mydata, Surv(OverallTime, Outcome))
head(km, 80)
 [1]  4.5+  7.8   7.1   7.9  10.6   6.9+  8.4+ 11.0   3.5   7.6   8.4   6.0 
[13]   NA   9.5  11.2  11.7   9.2   7.6?  4.1   4.7   9.7+  8.3+  6.0+  5.5+
[25]  6.4  11.4   3.8+ 10.2   3.0   6.4  11.3   6.5+  9.7   6.7   3.3+ 11.2+
[37]  7.8   7.0   6.3  10.2   7.0  11.2   9.7+  6.8   3.1   3.6   7.8   9.5+
[49]  6.0  10.4+ 11.2+  3.3+  7.4   9.2+  9.9  11.2+ 10.0   5.4   9.5   5.4 
[61]  5.9   8.4   4.1   9.2   7.3+  6.6   7.0+  8.6+  4.0   4.1  10.7   4.7 
[73]  6.9   6.6   5.3   8.0   9.3   8.4+  8.6+  8.8 
plot(km)

Kaplan-Meier Plot Log-Rank Test

# Drawing Survival Curves Using ggplot2
# https://rpkgs.datanovia.com/survminer/reference/ggsurvplot.html
dependentKM <- "Surv(OverallTime, Outcome)"
explanatoryKM <- "LVI"

mydata %>%
  finalfit::surv_plot(.data = .,
                      dependent = dependentKM,
                      explanatory = explanatoryKM,
                      xlab='Time (months)',
                      pval=TRUE,
                      legend = 'none',
                      break.time.by = 12,
                      xlim = c(0,60)
                      # legend.labs = c('a','b')
                      )

# Drawing Survival Curves Using ggplot2
# https://rpkgs.datanovia.com/survminer/reference/ggsurvplot.html

mydata %>%
  finalfit::surv_plot(.data = .,
                      dependent = "Surv(OverallTime, Outcome)",
                      explanatory = "LVI",
                      xlab='Time (months)',
                      pval=TRUE,
                      legend = 'none',
                      break.time.by = 12,
                      xlim = c(0,60)
                      # legend.labs = c('a','b')
                      )

5.3.2 Univariate Cox-Regression

library(finalfit)
library(survival)
explanatoryUni <- "LVI"
dependentUni <- "Surv(OverallTime, Outcome)"

tUni <- mydata %>% finalfit::finalfit(dependentUni, explanatoryUni)

knitr::kable(tUni[, 1:4], row.names = FALSE, align = c("l", "l", "r", "r", "r", "r"))
Dependent: Surv(OverallTime, Outcome) all HR (univariable)
LVI Absent 147 (100.0) -
Present 102 (100.0) 1.59 (1.15-2.20, p=0.005)
tUni_df <- tibble::as_tibble(tUni, .name_repair = "minimal") %>% janitor::clean_names()

tUni_df_descr <- paste0("When ", tUni_df$dependent_surv_overall_time_outcome[1], 
    " is ", tUni_df$x[2], ", there is ", tUni_df$hr_univariable[2], " times risk than ", 
    "when ", tUni_df$dependent_surv_overall_time_outcome[1], " is ", tUni_df$x[1], 
    ".")

When LVI is Present, there is 1.59 (1.15-2.20, p=0.005) times risk than when LVI is Absent.

5.3.3 Kaplan-Meier Median Survival

km_fit <- survfit(Surv(OverallTime, Outcome) ~ LVI, data = mydata)
km_fit
Call: survfit(formula = Surv(OverallTime, Outcome) ~ LVI, data = mydata)

   4 observations deleted due to missingness 
              n events median 0.95LCL 0.95UCL
LVI=Absent  144    100   22.0    14.3    31.0
LVI=Present 102     64   10.5     9.9    13.8
plot(km_fit)

# summary(km_fit)
km_fit_median_df <- summary(km_fit)
km_fit_median_df <- as.data.frame(km_fit_median_df$table) %>% janitor::clean_names() %>% 
    tibble::rownames_to_column()
km_fit_median_definition <- km_fit_median_df %>% dplyr::mutate(description = glue::glue("When {rowname}, median survival is {median} [{x0_95lcl} - {x0_95ucl}, 95% CI] months.")) %>% 
    dplyr::select(description) %>% dplyr::pull()

When LVI=Absent, median survival is 22 [14.3 - 31, 95% CI] months., When LVI=Present, median survival is 10.5 [9.9 - 13.8, 95% CI] months.

5.3.4 1-3-5-yr survival

summary(km_fit, times = c(12, 36, 60))
Call: survfit(formula = Surv(OverallTime, Outcome) ~ LVI, data = mydata)

4 observations deleted due to missingness 
                LVI=Absent 
 time n.risk n.event survival std.err lower 95% CI upper 95% CI
   12     75      52    0.617  0.0421        0.539        0.705
   36     19      35    0.252  0.0452        0.177        0.358

                LVI=Present 
 time n.risk n.event survival std.err lower 95% CI upper 95% CI
   12     23      49    0.383  0.0566       0.2870        0.512
   36      4      12    0.134  0.0488       0.0657        0.274
km_fit_summary <- summary(km_fit, times = c(12, 36, 60))

km_fit_df <- as.data.frame(km_fit_summary[c("strata", "time", "n.risk", "n.event", 
    "surv", "std.err", "lower", "upper")])
km_fit_definition <- km_fit_df %>% dplyr::mutate(description = glue::glue("When {strata}, {time} month survival is {scales::percent(surv)} [{scales::percent(lower)}-{scales::percent(upper)}, 95% CI].")) %>% 
    dplyr::select(description) %>% dplyr::pull()

When LVI=Absent, 12 month survival is 62% [54%-70.5%, 95% CI]., When LVI=Absent, 36 month survival is 25% [18%-35.8%, 95% CI]., When LVI=Present, 12 month survival is 38% [29%-51.2%, 95% CI]., When LVI=Present, 36 month survival is 13% [7%-27.4%, 95% CI].

source(here::here("R", "gc_survival.R"))

5.3.5 Survival Analysis LVI

Kaplan-Meier Plot Log-Rank Test

library(survival)
library(survminer)
library(finalfit)

mydata %>%
  finalfit::surv_plot('Surv(OverallTime, Outcome)', 'LVI', 
  xlab='Time (months)', pval=TRUE, legend = 'none',
    break.time.by = 12, xlim = c(0,60)

# legend.labs = c('a','b')

)

Univariate Cox-Regression

explanatoryUni <- "LVI"
dependentUni <- "Surv(OverallTime, Outcome)"
tUni <- mydata %>% finalfit(dependentUni, explanatoryUni, metrics = TRUE)

knitr::kable(tUni[, 1:4], row.names = FALSE, align = c("l", "l", "r", "r", "r", "r"))
Error in tUni[, 1:4]: incorrect number of dimensions

Univariate Cox-Regression Summary

tUni_df <- tibble::as_tibble(tUni, .name_repair = "minimal") %>% janitor::clean_names(dat = ., 
    case = "snake")


n_level <- dim(tUni_df)[1]

tUni_df_descr <- function(n) {
    paste0("When ", tUni_df$dependent_surv_overall_time_outcome[1], " is ", tUni_df$x[n + 
        1], ", there is ", tUni_df$hr_univariable[n + 1], " times risk than ", "when ", 
        tUni_df$dependent_surv_overall_time_outcome[1], " is ", tUni_df$x[1], ".")
    
}



results5 <- purrr::map(.x = c(2:n_level - 1), .f = tUni_df_descr)

print(unlist(results5))
[1] "When  is c(\"Absent\", \"Present\"), there is  times risk than when  is c(\"LVI\", \"\")."

Median Survival

km_fit <- survfit(Surv(OverallTime, Outcome) ~ LVI, data = mydata)
km_fit
Call: survfit(formula = Surv(OverallTime, Outcome) ~ LVI, data = mydata)

   4 observations deleted due to missingness 
              n events median 0.95LCL 0.95UCL
LVI=Absent  144    100   22.0    14.3    31.0
LVI=Present 102     64   10.5     9.9    13.8
km_fit_median_df <- summary(km_fit)
km_fit_median_df <- as.data.frame(km_fit_median_df$table) %>% janitor::clean_names(dat = ., 
    case = "snake") %>% tibble::rownames_to_column(.data = ., var = "LVI")



km_fit_median_definition <- km_fit_median_df %>% dplyr::mutate(description = glue::glue("When, LVI, {LVI}, median survival is {median} [{x0_95lcl} - {x0_95ucl}, 95% CI] months.")) %>% 
    dplyr::mutate(description = gsub(pattern = "thefactor=", replacement = " is ", 
        x = description)) %>% dplyr::select(description) %>% dplyr::pull()

km_fit_median_definition
When, LVI, LVI=Absent, median survival is 22 [14.3 - 31, 95% CI] months.
When, LVI, LVI=Present, median survival is 10.5 [9.9 - 13.8, 95% CI] months.

1-3-5-yr survival

summary(km_fit, times = c(12, 36, 60))
Call: survfit(formula = Surv(OverallTime, Outcome) ~ LVI, data = mydata)

4 observations deleted due to missingness 
                LVI=Absent 
 time n.risk n.event survival std.err lower 95% CI upper 95% CI
   12     75      52    0.617  0.0421        0.539        0.705
   36     19      35    0.252  0.0452        0.177        0.358

                LVI=Present 
 time n.risk n.event survival std.err lower 95% CI upper 95% CI
   12     23      49    0.383  0.0566       0.2870        0.512
   36      4      12    0.134  0.0488       0.0657        0.274
km_fit_summary <- summary(km_fit, times = c(12, 36, 60))

km_fit_df <- as.data.frame(km_fit_summary[c("strata", "time", "n.risk", "n.event", 
    "surv", "std.err", "lower", "upper")])

km_fit_df
       strata time n.risk n.event      surv    std.err      lower     upper
1  LVI=Absent   12     75      52 0.6165782 0.04211739 0.53931696 0.7049078
2  LVI=Absent   36     19      35 0.2520087 0.04515881 0.17737163 0.3580528
3 LVI=Present   12     23      49 0.3833784 0.05662684 0.28701265 0.5120993
4 LVI=Present   36      4      12 0.1340646 0.04881983 0.06566707 0.2737036
km_fit_definition <- km_fit_df %>% dplyr::mutate(description = glue::glue("When {strata}, {time} month survival is {scales::percent(surv)} [{scales::percent(lower)}-{scales::percent(upper)}, 95% CI].")) %>% 
    dplyr::select(description) %>% dplyr::pull()

km_fit_definition
When LVI=Absent, 12 month survival is 62% [54%-70.5%, 95% CI].
When LVI=Absent, 36 month survival is 25% [18%-35.8%, 95% CI].
When LVI=Present, 12 month survival is 38% [29%-51.2%, 95% CI].
When LVI=Present, 36 month survival is 13% [7%-27.4%, 95% CI].
summary(km_fit)$table
            records n.max n.start events   *rmean *se(rmean) median 0.95LCL
LVI=Absent      144   144     144    100 24.71341   1.571856   22.0    14.3
LVI=Present     102   102     102     64 17.48672   1.904576   10.5     9.9
            0.95UCL
LVI=Absent     31.0
LVI=Present    13.8
km_fit_median_df <- summary(km_fit)
results1html <- as.data.frame(km_fit_median_df$table) %>% janitor::clean_names(dat = ., 
    case = "snake") %>% tibble::rownames_to_column(.data = ., var = "LVI")

results1html[, 1] <- gsub(pattern = "thefactor=", replacement = "", x = results1html[, 
    1])

knitr::kable(results1html, row.names = FALSE, align = c("l", rep("r", 9)), format = "html", 
    digits = 1)
LVI records n_max n_start events rmean se_rmean median x0_95lcl x0_95ucl
LVI=Absent 144 144 144 100 24.7 1.6 22.0 14.3 31.0
LVI=Present 102 102 102 64 17.5 1.9 10.5 9.9 13.8

Pairwise Comparisons

5.3.6 Pairwise comparison

dependentKM <- "Surv(OverallTime, Outcome)"
explanatoryKM <- "TStage"

mydata %>%
  finalfit::surv_plot(.data = .,
                      dependent = dependentKM,
                      explanatory = explanatoryKM,
                      xlab='Time (months)',
                      pval=TRUE,
                      legend = 'none',
                      break.time.by = 12,
                      xlim = c(0,60)
                      # legend.labs = c('a','b')
                      )

km_fit
Call: survfit(formula = Surv(OverallTime, Outcome) ~ LVI, data = mydata)

   4 observations deleted due to missingness 
              n events median 0.95LCL 0.95UCL
LVI=Absent  144    100   22.0    14.3    31.0
LVI=Present 102     64   10.5     9.9    13.8
print(km_fit, 
      scale=1,
      digits = max(options()$digits - 4,3),
      print.rmean=getOption("survfit.print.rmean"),
      rmean = getOption('survfit.rmean'),
      print.median=getOption("survfit.print.median"),
      median = getOption('survfit.median')
      
      )
Call: survfit(formula = Surv(OverallTime, Outcome) ~ LVI, data = mydata)

   4 observations deleted due to missingness 
              n events median 0.95LCL 0.95UCL
LVI=Absent  144    100   22.0    14.3    31.0
LVI=Present 102     64   10.5     9.9    13.8

5.3.7 Multivariate Analysis Survival

explanatory = c(“age.factor”, “sex.factor”, “obstruct.factor”, “perfor.factor”) dependent = ‘mort_5yr’ colon_s %>% hr_plot(dependent, explanatory)

library(survival)
library(survminer)
library(finalfit)
mb_followup %>%
  finalfit::surv_plot('Surv(OverallTime, Outcome)', 'Operation', 
  xlab='Time (months)', pval=TRUE, legend = 'none',
  # pval.coord
    break.time.by = 12, xlim = c(0,60), ylim = c(0.8, 1)

# legend.labs = c('a','b')

)

Univariate Cox-Regression

explanatoryUni <- "Operation"
dependentUni <- "Surv(OverallTime, Outcome)"
tUni <- mb_followup %>% finalfit(dependentUni, explanatoryUni)

knitr::kable(tUni[, 1:4], row.names = FALSE, align = c("l", "l", "r", "r", "r", "r"))

Univariate Cox-Regression Summary

tUni_df <- tibble::as_tibble(tUni, .name_repair = "minimal") %>% janitor::clean_names(dat = ., 
    case = "snake")


n_level <- dim(tUni_df)[1]

tUni_df_descr <- function(n) {
    paste0("When ", tUni_df$dependent_surv_overall_time_outcome[1], " is ", tUni_df$x[n + 
        1], ", there is ", tUni_df$hr_univariable[n + 1], " times risk than ", "when ", 
        tUni_df$dependent_surv_overall_time_outcome[1], " is ", tUni_df$x[1], ".")
    
}



results5 <- purrr::map(.x = c(2:n_level - 1), .f = tUni_df_descr)

print(unlist(results5))

Median Survival

km_fit <- survfit(Surv(OverallTime, Outcome) ~ Operation, data = mb_followup)

# km_fit

# summary(km_fit)

km_fit_median_df <- summary(km_fit)
km_fit_median_df <- as.data.frame(km_fit_median_df$table) %>% janitor::clean_names(dat = ., 
    case = "snake") %>% tibble::rownames_to_column(.data = ., var = "Derece")

km_fit_median_df

# km_fit_median_df %>% knitr::kable(format = 'latex') %>%
# kableExtra::kable_styling(latex_options='scale_down')

km_fit_median_definition <- km_fit_median_df %>% dplyr::mutate(description = glue::glue("When, Derece, {Derece}, median survival is {median} [{x0_95lcl} - {x0_95ucl}, 95% CI] months.")) %>% 
    dplyr::mutate(description = gsub(pattern = "thefactor=", replacement = " is ", 
        x = description)) %>% dplyr::select(description) %>% dplyr::pull()

# km_fit_median_definition

1-3-5-yr survival

summary(km_fit, times = c(12, 36, 60))

km_fit_summary <- summary(km_fit, times = c(12, 36, 60))

km_fit_df <- as.data.frame(km_fit_summary[c("strata", "time", "n.risk", "n.event", 
    "surv", "std.err", "lower", "upper")])

km_fit_df %>% knitr::kable(format = "latex") %>% kableExtra::kable_styling(latex_options = "scale_down")





km_fit_definition <- km_fit_df %>% dplyr::mutate(description = glue::glue("When {strata}, {time} month survival is {scales::percent(surv)} [{scales::percent(lower)}-{scales::percent(upper)}, 95% CI].")) %>% 
    dplyr::select(description) %>% dplyr::pull()

km_fit_definition

Pairwise Comparisons

survminer::pairwise_survdiff(formula = Surv(OverallTime, Outcome) ~ Operation, data = mb_followup, 
    p.adjust.method = "BH")

library(gt)
library(gtsummary)

library(survival)
fit1 <- survfit(Surv(ttdeath, death) ~ trt, trial)
tbl_strata_ex1 <- tbl_survival(fit1, times = c(12, 24), label = "{time} Months")

fit2 <- survfit(Surv(ttdeath, death) ~ 1, trial)
tbl_nostrata_ex2 <- tbl_survival(fit2, probs = c(0.1, 0.2, 0.5), header_estimate = "**Months**")

6 Interactive Survival Analysis

Codes for generating Survival Analysis.33

Codes for generating Shiny Survival Analysis.34


7 Correlation

7.1 Correlation Analysis

Codes for generating correlation analysis.35

https://stat.ethz.ch/R-manual/R-patched/library/stats/html/cor.test.html

https://neuropsychology.github.io/psycho.R/2018/05/20/correlation.html

devtools::install_github("neuropsychology/psycho.R")  # Install the newest version

remove.packages("psycho")
renv::install("neuropsychology/psycho.R@0.4.0")
# devtools::install_github("neuropsychology/psycho.R@0.4.0")

library(psycho)
<!-- library(tidyverse) -->

cor <- psycho::affective %>% 
  correlation()

summary(cor)


plot(cor)


print(cor)
summary(cor) %>% 
  knitr::kable(format = "latex") %>% 
  kableExtra::kable_styling(latex_options="scale_down")


ggplot(mydata, aes(x = tx_zamani_verici_yasi, y = trombosit)) +
  geom_point() + 
  geom_smooth(method = lm, size = 1)

8 Models

Codes used in models36
Use these descriptions to add autoreporting of new models

8.0.1 Linear Model

generate automatic reporting of model via easystats/report 📦

library(report)
model <- lm(Sepal.Length ~ Species, data = iris)
report::report(model)
We fitted a linear model (estimated using OLS) to predict Sepal.Length with Species (formula = Sepal.Length ~ Species). Standardized parameters were obtained by fitting the model on a standardized version of the dataset. Effect sizes were labelled following Funder's (2019) recommendations.

The model explains a significant and substantial proportion of variance (R2 = 0.62, F(2, 147) = 119.26, p < .001, adj. R2 = 0.61). The model's intercept, corresponding to Sepal.Length = 0 and Species = setosa, is at 5.01 (SE = 0.07, 95% CI [4.86, 5.15], p < .001). Within this model:

  - The effect of Speciesversicolor is positive and can be considered as very large and significant (beta = 1.12, SE = 0.12, 95% CI [0.88, 1.37], std. beta = 1.12, p < .001).
  - The effect of Speciesvirginica is positive and can be considered as very large and significant (beta = 1.91, SE = 0.12, 95% CI [1.66, 2.16], std. beta = 1.91, p < .001).

Table report for a linear model

model <- lm(Sepal.Length ~ Petal.Length + Species, data=iris)
r <- report(model)
to_text(r)
to_table(r)

8.0.2 General Linear Models (GLMs)

https://stat.ethz.ch/R-manual/R-devel/library/stats/html/glm.html

model <- glm(vs ~ mpg + cyl, data=mtcars, family="binomial")
r <- report(model)

to_fulltext(r)
to_fulltable(r)
Where a multivariable model contains a subset of the variables specified in the full univariable set, this can be specified.

explanatory = c("age.factor", "sex.factor", "obstruct.factor", "perfor.factor")
explanatory.multi = c("age.factor", "obstruct.factor")
dependent = 'mort_5yr'
colon_s %>%
  summarizer(dependent, explanatory, explanatory.multi)

Random effects.

e.g. lme4::glmer(dependent ~ explanatory + (1 | random_effect), family="binomial")

explanatory = c("age.factor", "sex.factor", "obstruct.factor", "perfor.factor")
explanatory.multi = c("age.factor", "obstruct.factor")
random.effect = "hospital"
dependent = 'mort_5yr'
colon_s %>%
  summarizer(dependent, explanatory, explanatory.multi, random.effect)

metrics=TRUE provides common model metrics.

colon_s %>%
  summarizer(dependent, explanatory, explanatory.multi,  metrics=TRUE)

Cox proportional hazards

e.g. survival::coxph(dependent ~ explanatory)

explanatory = c("age.factor", "sex.factor", "obstruct.factor", "perfor.factor")
dependent = "Surv(time, status)"

colon_s %>%
    summarizer(dependent, explanatory)

Rather than going all-in-one, any number of subset models can be manually added on to a summary.factorlist() table using summarizer.merge(). This is particularly useful when models take a long-time to run or are complicated.

Note requirement for glm.id=TRUE. fit2df is a subfunction extracting most common models to a dataframe.

explanatory = c("age.factor", "sex.factor", "obstruct.factor", "perfor.factor")
explanatory.multi = c("age.factor", "obstruct.factor")
random.effect = "hospital"
dependent = 'mort_5yr'

# Separate tables
colon_s %>%
  summary.factorlist(dependent, explanatory, glm.id=TRUE) -> example.summary

colon_s %>%
  glmuni(dependent, explanatory) %>%
  fit2df(estimate.suffix=" (univariable)") -> example.univariable

colon_s %>%
  glmmulti(dependent, explanatory) %>%
  fit2df(estimate.suffix=" (multivariable)") -> example.multivariable


colon_s %>%
  glmmixed(dependent, explanatory, random.effect) %>%
  fit2df(estimate.suffix=" (multilevel") -> example.multilevel

# Pipe together
example.summary %>%
  summarizer.merge(example.univariable) %>%
  summarizer.merge(example.multivariable) %>%
  summarizer.merge(example.multilevel) %>%
  select(-c(glm.id, index)) -> example.final
example.final

Cox Proportional Hazards example with separate tables merged together.

explanatory = c("age.factor", "sex.factor", "obstruct.factor", "perfor.factor")
explanatory.multi = c("age.factor", "obstruct.factor")
dependent = "Surv(time, status)"

# Separate tables
colon_s %>%
    summary.factorlist(dependent, explanatory, glm.id=TRUE) -> example2.summary

colon_s %>%
    coxphuni(dependent, explanatory) %>%
    fit2df(estimate.suffix=" (univariable)") -> example2.univariable

colon_s %>%
  coxphmulti(dependent, explanatory.multi) %>%
  fit2df(estimate.suffix=" (multivariable)") -> example2.multivariable

# Pipe together
example2.summary %>%
    summarizer.merge(example2.univariable) %>%
    summarizer.merge(example2.multivariable) %>%
    select(-c(glm.id, index)) -> example2.final
example2.final







# OR plot
explanatory = c("age.factor", "sex.factor", "obstruct.factor", "perfor.factor")
dependent = 'mort_5yr'
colon_s %>%
  or.plot(dependent, explanatory)
# Previously fitted models (`glmmulti()` or `glmmixed()`) can be provided directly to `glmfit`

# HR plot (not fully tested)
explanatory = c("age.factor", "sex.factor", "obstruct.factor", "perfor.factor")
dependent = "Surv(time, status)"
colon_s %>%
  hr.plot(dependent, explanatory, dependent_label = "Survival")
# Previously fitted models (`coxphmulti`) can be provided directly using `coxfit`

8.0.3 ANOVA

8.0.4 Bayesian

# Full report for a Bayesian logistic mixed model with effect sizes
library(rstanarm)

stan_glmer(vs ~ mpg + (1|cyl), data=mtcars, family="binomial") %>% 
  report(standardize="smart", effsize="cohen1988") %>% 
  to_fulltext()

8.0.5 lme4: Mixed-effects models in R

https://github.com/lme4/lme4/

8.0.6 indices of model quality and goodness of fit

Test if your model is a good model

https://easystats.github.io/performance/


Some Text ile sağkalım açısından bir ilişki bulunmamıştır (p = 0.22).

my_text <- kableExtra::text_spec("Some Text", color = "red", background = "yellow")
# `r my_text`

Text Here





Since R Markdown use the bootstrap framework under the hood. It is possible to benefit its powerful grid system. Basically, you can consider that your row is divided in 12 subunits of same width. You can then choose to use only a few of this subunits.



Here, I use 3 subunits of size 4 (4x3=12). The last column is used for a plot. You can read more about the grid system here. I got this result showing the following code in my R Markdown document.


9 Tabs for sub-chapters

9.1 First

content of sub-chapter #1

9.2 Second

content of sub-chapter #2

9.3 Third

content of sub-chapter #3

Block rmdnote

Block rmdtip

Block warning


10 Discussion

  • Interpret the results in context of the working hypothesis elaborated in the introduction and other relevant studies; include a discussion of limitations of the study.

  • Discuss potential clinical applications and implications for future research

References

Knijn, N., F. Simmer, and I. D. Nagtegaal. 2015. “Recommendations for Reporting Histopathology Studies: A Proposal.” Virchows Archiv 466 (6): 611–15. https://doi.org/10.1007/s00428-015-1762-3.

Schmidt, Robert L., Deborah J. Chute, Jorie M. Colbert-Getz, Adolfo Firpo-Betancourt, Daniel S. James, Julie K. Karp, Douglas C. Miller, et al. 2017. “Statistical Literacy Among Academic Pathologists: A Survey Study to Gauge Knowledge of Frequently Used Statistical Tests Among Trainees and Faculty.” Archives of Pathology & Laboratory Medicine 141 (2): 279–87. https://doi.org/10.5858/arpa.2016-0200-OA.


  1. From Table 1: Proposed items for reporting histopathology studies. Recommendations for reporting histopathology studies: a proposal Virchows Arch (2015) 466:611–615 DOI 10.1007/s00428-015-1762-3 https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4460276/↩︎

  2. From Table 1: Proposed items for reporting histopathology studies. Recommendations for reporting histopathology studies: a proposal Virchows Arch (2015) 466:611–615 DOI 10.1007/s00428-015-1762-3 https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4460276/↩︎

  3. See childRmd/_01header.Rmd file for other general settings↩︎

  4. Change echo = FALSE to hide codes after knitting and Change cache = TRUE to knit quickly. Change error=TRUE to continue rendering while errors are present.↩︎

  5. See childRmd/_02fakeData.Rmd file for other codes↩︎

  6. Synthea The validity of synthetic clinical data: a validation study of a leading synthetic data generator (Synthea) using clinical quality measures. BMC Med Inform Decis Mak 19, 44 (2019) doi:10.1186/s12911-019-0793-0↩︎

  7. https://bmcmedinformdecismak.biomedcentral.com/articles/10.1186/s12911-019-0793-0↩︎

  8. Synthetic Patient Generation↩︎

  9. Basic Setup and Running↩︎

  10. intelligent patient data generator (iPDG)↩︎

  11. https://medium.com/free-code-camp/how-our-test-data-generator-makes-fake-data-look-real-ace01c5bde4a↩︎

  12. https://forums.librehealth.io/t/demo-data-generation/203↩︎

  13. https://mihin.org/services/patient-generator/↩︎

  14. lung, cancer, breast datası ile birleştir↩︎

  15. See childRmd/_03importData.Rmd file for other codes↩︎

  16. See childRmd/_04briefSummary.Rmd file for other codes↩︎

  17. https://www.hhs.gov/hipaa/index.html↩︎

  18. Kişisel verilerin kaydedilmesi ve kişisel verileri hukuka aykırı olarak verme veya ele geçirme Türk Ceza Kanunu’nun 135. ve 136. maddesi kapsamında bizim hukuk sistemimizde suç olarak tanımlanmıştır. Kişisel verilerin kaydedilmesi suçunun cezası 1 ila 3 yıl hapis cezasıdır. Suçun nitelikli hali ise, kamu görevlisi tarafından görevin verdiği yetkinin kötüye kullanılarak veya belirli bir meslek veya sanatın sağladığı kolaylıktan yararlanılarak işlenmesidir ki bu durumda suçun cezası 1.5 ile 4.5 yıl hapis cezası olacaktır.↩︎

  19. See childRmd/_06variableTypes.Rmd file for other codes↩︎

  20. See childRmd/_07overView.Rmd file for other codes↩︎

  21. Statistical Literacy Among Academic Pathologists: A Survey Study to Gauge Knowledge of Frequently Used Statistical Tests Among Trainees and Faculty. Archives of Pathology & Laboratory Medicine: February 2017, Vol. 141, No. 2, pp. 279-287. https://doi.org/10.5858/arpa.2016-0200-OA↩︎

  22. From Table 1: Proposed items for reporting histopathology studies. Recommendations for reporting histopathology studies: a proposal Virchows Arch (2015) 466:611–615 DOI 10.1007/s00428-015-1762-3 https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4460276/↩︎

  23. See childRmd/_08dataDictionary.Rmd file for other codes↩︎

  24. See childRmd/_09cleanRecode.Rmd file for other codes↩︎

  25. See childRmd/_10impute.Rmd file for other codes↩︎

  26. See childRmd/_11descriptives.Rmd file for other codes↩︎

  27. See childRmd/_12crossTables.Rmd file for other codes↩︎

  28. See childRmd/_13plots.Rmd file for other codes↩︎

  29. See childRmd/_14pairedTests.Rmd file for other codes↩︎

  30. See childRmd/_15hypothesisTests.Rmd file for other codes↩︎

  31. See childRmd/_18survival.Rmd file for other codes, and childRmd/_19shinySurvival.Rmd for shiny application↩︎

  32. JAMA retraction after miscoding – new Finalfit function to check recoding↩︎

  33. See childRmd/_18survival.Rmd file for other codes↩︎

  34. See childRmd/_19shinySurvival.Rmd file for other codes↩︎

  35. See childRmd/_20correlation.Rmd file for other codes↩︎

  36. See childRmd/_21models.Rmd file for other codes↩︎

  37. See childRmd/_23footer.Rmd file for other codes↩︎

  38. Smith AM, Katz DS, Niemeyer KE, FORCE11 Software Citation Working Group. (2016) Software Citation Principles. PeerJ Computer Science 2:e86. DOI: 10.7717/peerj-cs.86 https://www.force11.org/software-citation-principles↩︎

 

A work by Serdar Balci

drserdarbalci@gmail.com