JoeToolFunctions.R 5.26 KB
Newer Older
Jalbert, Joe's avatar
Jalbert, Joe committed
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
#' Standard Error
#'
#' @param data The vector or dataset over which the standard error is calculated
#'
#' @return The calulated Standard Error
#' @importFrom stats sd
#' @export
#'
#' @examples
#' x=c(10, 20, 30, 40, 50)
#' se(x)
se=function(data){
  return(sd(data, na.rm=T)/sqrt(sum(!is.na(data))))
}

#' Flexible Workbook Creation Function
#'
#' @param df The dataframe to be saved
#' @param filename Name of the file, including the xlsx extension
#' @param sheetBy Option to create sheets based on the values in a particular column of the dataframe
#' @param keepNames The function automatically renames the column names with \code{\link[stringr:case]{stringr::str_to_sentence()}}. Setting this to TRUE will leave your column names unaltered
#'
#' @return Creates a Excel workbook from the supplied dataframe
#' @import tidyverse
#' @import xlsx
#' @import lazyeval
#' @export
#'
#' @examples
#' library(datasets)
#' library(tidyverse)
#' library(xlsx)
#' data(iris)
#' wbsave(iris, "Iris report.xlsx")
#' wbsave(iris, "Iris report - by Species.xlsx", sheetBy="Species", keepNames=TRUE)
wbsave=function(df,filename,sheetBy=NULL,keepNames=FALSE){
  if(keepNames==FALSE) {names(df)=str_to_sentence(names(df))}
  if(!is.null(sheetBy)) {sheetBy=str_to_sentence(sheetBy)}
  wb=createWorkbook()
  TABLE_COLNAMES_STYLE <- CellStyle(wb)+ Font(wb,  heightInPoints=11, color="#44546A", isBold=TRUE)+
    Border(color = "#8EA9DB", position = "BOTTOM", pen="BORDER_THICK")
  sheetAll=createSheet(wb,sheetName = "All")
  wbdata=NULL
  wbAll=NULL
  if(!is.null(sheetBy)) {l=levels(as.factor(df[[sheetBy]]))}
  if(!is.null(sheetBy)) {for (i in l){
    filter_criteria <- interp(~y == x, .values=list(y = as.name(sheetBy), x = i))
    wbdata=df%>%
      filter_(filter_criteria)
    wbAll[[i]]=wbdata
    sheet=createSheet(wb,sheetName = i)
    addDataFrame(wbdata,sheet, colnamesStyle = TABLE_COLNAMES_STYLE)
    autoSizeColumn(sheet, colIndex = 1:ncol(df)+1)
  }}
  addDataFrame(df,sheetAll, colnamesStyle = TABLE_COLNAMES_STYLE)
  autoSizeColumn(sheetAll, colIndex = 1:ncol(df)+1)
  saveWorkbook(wb, filename)
}

#' Creates a Report with t-tests for a Vector of Outcomes
#'
#' @param df Dataframe to use for the report
#' @param Measures A vector of string names for columns to take means for. Order will be used to make the report.
#' @param Factor A binary factor which will be used for comparisons.  Must be 1 (Treatment) or 0 (No_Treatment)
#' @param paired Options for running t-test.  "Yes" forces paired, t-tests, "No" assumes no pairing, and "Try" will try a paired t-test, and follow up with a unpaired t-test if it fails to run.
#'
#' @return
#' @import tidyverse
#' @export
#'
#'
#' @examples

report=function(df, Measures, Factor, paired=c("Yes", "No", "Try")){
  output=df%>%
    group_by(.data[[Factor]])%>%
    summarise_at(vars(Measures), mean,na.rm=T)%>%
    mutate(Levels =ifelse(.data[[Factor]]==1, "Treatment", "No_Treatment"))%>%
    select(-.data[[Factor]])%>%
    pivot_longer(cols=Outcomes, names_to = "Measure")%>%
    pivot_wider(names_from = "Levels", values_from = "value")%>%
    mutate(Diff=Treatment-No_Treatment)%>%
    mutate(Measure=factor(.data$Measure,levels=Measures))%>%
    arrange(Measure)


  reportpvaluecount=0
Jalbert, Joe's avatar
Jalbert, Joe committed
88
  p=NULL
Jalbert, Joe's avatar
Jalbert, Joe committed
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
  for(i in Measures){
    reportpvaluecount=reportpvaluecount+1
    form=paste0(i, "~",Factor)
    if(paired=="Yes"){p[reportpvaluecount]=as.numeric(try(t.test(formula=as.formula(form), data=df, paired=T)$p.value))}
    else if (paired=="No"){p[reportpvaluecount]=as.numeric(try(t.test(formula=as.formula(form), data=df)$p.value))}
    else {p[reportpvaluecount]=tryCatch(as.numeric(t.test(formula=as.formula(form), data=df, paired=T)$p.value), error=function(err)
      p[reportpvaluecount]=as.numeric(try(t.test(formula=as.formula(form), data=df)$p.value)), finally = NA)}

  }

  output$p_value=p

  output=output%>%
    mutate(Signif=ifelse(p_value<.001, "***",ifelse(p_value<.01, "**",
                                                    ifelse(p_value<.05, "*", ifelse(p_value<.1, ".", "")))))
}
Jalbert, Joe's avatar
Jalbert, Joe committed
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137


#' Appends Interactions to a \code{\link[JoeTools:report]{report()}} Output Dataframe
#'
#' @param report The report dataframe that is being appended
#' @param df The orginal dataframe used for the report
#' @param Measures A vector of string names for the dependant variable. Order must be the same as the repor
#' @param Factor A binary factor (as a string) which will be used for comparisons.  Must be 1 (Treatment) or 0 (No_Treatment)
#' @param Interaction A factor column name (as a string) that will be used as an interacting independant variable.
#'
#' @return
#' @import tidyverse
#' @export
#'
#' @examples
appendInteraction=function(report,df, Measures,Factor,Interaction){
  p=NULL
  reportpvaluecount=0
  for(i in Measures){
    reportpvaluecount=reportpvaluecount+1
    form=paste0(i, "~",Factor,"*",Interaction)
    p[reportpvaluecount]=tryCatch(summary(aov(formula = as.formula(form), data=df))[[1]][["Pr(>F)"]][[3]], error=function(err) NA)
  }
  name=paste0(Interaction,"_pvalue")
  name2=paste0(Interaction,"_Int")
  output=report
  output[[name]]=p
  output=output%>%
    mutate(NewP=ifelse(.data[[name]]<.001, "***",ifelse(.data[[name]]<.01, "**",
                                                        ifelse(.data[[name]]<.05, "*", ifelse(.data[[name]]<.1, ".", "")))))%>%
    rename(!!name2:=NewP)
  
}