R/Self.R
In sebastiz/SelfAwaRe:
Defines functions
ProcDiagnosisUnit
ReportOP
library(stringr)
library(ggplot2)
library(lattice)
library(compare)
library(XLConnect)
library(grid)
library(reshape2)
library(gtools)
library(gplots)
library("tm")
library("SnowballC")
library("wordcloud")
library("RColorBrewer")
library(splitstackshape)
library(knitr)
library(xtable)
library(pander)
library(data.table)
library(gridExtra)
library(pheatmap)
library(tidyr)
library(EndoMineR)
# local({
# hook_plot = knit_hooks$get('plot')
# knit_hooks$set(plot = function(x, options) {
# x = paste(x, collapse = '.')
# if (!grepl('\\.svg', x)) return(hook_plot(x, options))
# # read the content of the svg image and write it out without <?xml ... ?>
# paste(readLines(x)[-1], collapse = '\n')
# #paste("<figure><img src=\"", opts_knit$get("base.url"), paste(x, collapse = "."),
# # "\"><figcaption>", options$fig.cap, "</figcaption></figure>", sep = "")
# })
# })
#For the imag extraction- need to get this off EndoSoft. Choose the query called Images Captured with Proc Data Audit for the month beginning 1 to the end of the month
#Then export HTML to folder called "S:\Gastroenterology\Seb\R\Data\SelfAudit" with styles and Fixed width and JPEG (default settings selected) and it will create a folder called Images Captured with Proc Data Audit.files. The html will be one level up
############################################# ############ DATA ACQUISITION ################## ############################################# #############################################
Self = loadWorkbook("S:\\Gastroenterology\\Seb\\R\\Data\\SelfAudit\\Gastroscopy_TCR2232HistoReport_run%2004-01-2017(Apr18).xlsx")
Self = XLConnect::readWorksheet(Self, sheet="Data",header=TRUE)
############################################# ############ DATA CLEANING ################## ############################################# #############################################
names(Self)<-c("PatientID", "NHSno", "PatientName", "birthdaynum","birthmonthnum" ,"birthyearnum" ,"Endo_ResultName", "Endo_ResultPerformed", "Endo_ResultEntered", "Endo_ResultText", "Histo_ResultName", "Histo_ResultPerformed", "Histo_ResultEntered", "Histo_ResultText")
#Remove the first row if necessary
Self<-Self[2:nrow(Self),]
#Get rid of everything without an endoscopy report (because pre 2007 there won't be one, just histopath)
Self<-subset(Self,!(is.na(Self["Endo_ResultName"])))
#Split the spreadsheet up into new columns and then get rid of the original ones
#For the Endoscopy text
source("S:\\Gastroenterology\\Seb\\R\\Scripts\\Generics\\CleanUp.R")
#Self<-EndoscChopper(Self)
Self$Endo_ResultText<-gsub("2nd Endoscopist:","Second endoscopist:",Self$Endo_ResultText)
EndoscTree<-list("Patient Name:","Date of Birth:","General Practicioner:","Hospital Number:",
"Date of Procedure:","Endoscopist:","Second endoscopist:","Trainee:","Referring Physician:","Nurses:","Medications",
"Instrument","Extent of Exam:","INDICATIONS FOR EXAMINATION","PROCEDURE PERFORMED","FINDINGS",
"ENDOSCOPIC DIAGNOSIS","RECOMMENDATIONS","COMMENTS","FOLLOW UP","")
for(i in 1:(length(EndoscTree)-1)) {
Self<-Extractor(Self,"Endo_ResultText",as.character(EndoscTree[i]),
as.character(EndoscTree[i+1]),as.character(EndoscTree[i]))
}
#Extra tidying...DATES ARE A PAIN IN THE ASS....
Self$PROCEDUREPERFORMED<-gsub("(Withdrawal|Quality|ENDOSCOPIC).*","",Self$PROCEDUREPERFORMED)
Self$PROCEDUREPERFORMED<-gsub("\\s+[(]OGD[)]","",Self$PROCEDUREPERFORMED)
#Convert date time to date in R
Self$Endo_ResultEntered<-as.Date(as.POSIXct(Self$Endo_ResultEntered,format="%d/%m/%Y"))
#And sometimes this is the one to use...
#Self$Endo_ResultEntered<-as.Date(format(Self$Endo_ResultEntered), "%d/%m/%Y")
#Change the names of the columns as EPR keep on changing things around:
names(Self)<-c("PatientID","NHSno","PatientName","birthdaynum","birthmonthnum","birthyearnum","Endo_ResultName","Endo_ResultPerformed","Endo_ResultEntered","Endo_ResultText","Histo_ResultName","Histo_ResultPerformed","Histo_ResultEntered","Histo_ResultText","DateofBirth","GeneralPracticioner","HospitalNumber","DateofProcedure","Endo_Endoscopist","SecondEndo_Endoscopist","Trainee","ReferringPhysician","Nurses","Medications","Instrument","ExtentofExam","IndicationsFroExamination","ProcPerformed","Findings","EndoscopDiagn","Recommend","COMMENTS","FollowUp")
Self<-HistolChopper(Self)
Self$FindingsAlgoNegs<-str_extract_all(Self$Findings,".*[Nn]egative.*|.*[Nn]either.*|[Nn]o .*?[\\.\n]|.*[Nn]ormal.*?[\\.\n]|There (?:is|was|are) no.*?\\.|.*None.*?[\\.\\n]|.*unremarkable?\\.|^Negative.*?[\\.\\n]|[Oo]therwise (?:normal|unremarkable).*?[\\.\\n]")
Self$EndoscopDiagnAlgoNegs<-str_extract_all(Self$EndoscopDiagn,".*[Nn]egative.*|.*[Nn]either.*|[Nn]o .*?[\\.\\n]|.*[Nn]ormal.*?[\\.\\n]|There (?:is|was|are) no.*?\\.|.*None.*?[\\.\\n]|.*unremarkable?\\.|^Negative.*?[\\.\\n]|[Oo]therwise (?:normal|unremarkable).*?[\\.\\n]")
#Get histopath into a proper format for line by line extraction:
Self$Histo_ResultTextForFindings<-gsub("\\.\\.","\\.",Self$Histo_ResultTextForFindings)
Self$Histo_ResultTextForFindings<-gsub("\\.","\\.\n",Self$Histo_ResultTextForFindings)
Self$Histo_ResultTextForFindings<-gsub(":","",Self$Histo_ResultTextForFindings)
Self$Histo_ResultTextForFindingsAlgoNegs<-str_extract_all(Self$Histo_ResultTextForFindings,".*[Nn]egative.*|.*[Nn]either.*|[Nn]o .*?[\\.\\n]|.*[Nn]ormal.*?[\\.\\n]|There (?:is|was|are) no.*?[\\.\\n]|\\..*None.*?[\\.\\n]|\\..*unremarkable.*?[\\.\\n]|^Negative.*?[\\.\\n]|[Oo]therwise (?:normal|unremarkable).*?[\\.\\n]")
SelfWithNegs<-data.frame(Self["PatientID"],Self["Findings"],Self["FindingsAlgoNegs"],Self["EndoscopDiagn"],Self["EndoscopDiagnAlgoNegs"],Self["Histo_ResultTextForFindings"],Self["Histo_ResultTextForFindingsAlgoNegs"])
#Import the image file here from html report from Endosoft:
#You need to go to Endosoft, then pick the query under the GRS tab, under GSTT, called Images Captured with Proc Data Audit.files
#Then run for the month you are interested in
#Then export as HTML into the folder called "S:\Gastroenterology\Seb\R\Data\SelfAudit" It will name the folder itself as Images Captured with Proc Data Audit.files and the html file will be in the SelfAudit level
#Then run the script
############################### Drop in the image table here############################### ###############################
source("S:\\Gastroenterology\\Seb\\R\\Scripts\\SelfAudit\\SelfimgExtractionScripts.R")
images<-MyImgLibrary()
source("S:\\Gastroenterology\\Seb\\R\\Scripts\\Generics\\CleanUp.R")
Self<-merge(images,Self,by=c("Endo_ResultEntered","PatientID"),all.y=T)
############################################# ######## DATA CLEANING ###################### #############################################
#Tidy the whole thing up:
Self$Endo_Endoscopist.x<-NULL
names(Self)[names(Self) == 'Endo_Endoscopist.y'] <-"Endo_Endoscopist"
Self$V1.x<-NULL
Self$V1<-NULL
rmarkdown::render("S:\\Gastroenterology\\Seb\\R\\Scripts\\SelfAudit\\Self2SANDBOX.Rmd",output_file=paste("MonthlyEndoscopyUnitReport2",Sys.Date(),'.docx'))
#Count the number of sentences in a column:
#Get rid of the starting period
Self$Findings<-gsub(".\n.\n",".\n",Self$Findings)
Self$TotalLines<-str_count(Self$Findings, ".\n")
Self$TP<-str_count(Self$FindingsAlgoNegs, "\\.")
Self$TN<-Self$TotalLines-Self$TP
Self$FP<-0
Self$FN<-0
#Self$Histo_ResultTextForFindings2<-gsub("\\.",".\n",Self$Histo_ResultTextForFindings)
Self$TotalLinesHistoResultTextForFindings<-str_count(Self$Histo_ResultTextForFindings, ".\n")
Self$TPHistoResultTextForFindings<-str_count(Self$Histo_ResultTextForFindingsAlgoNegs, "\\.")
Self$TNHistoResultTextForFindings<-Self$TotalLinesHistoResultTextForFindings-Self$TPHistoResultTextForFindings
Self$FPHistoResultTextForFindings<-0
Self$FNHistoResultTextForFindings<-0
SelfForSheet<-data.frame(Self["Histo_ResultTextForFindings"],Self["Histo_ResultTextForFindingsAlgoNegs"],Self["TotalLinesHistoResultTextForFindings"],Self["TPHistoResultTextForFindings"],Self["FPHistoResultTextForFindings"],Self["TNHistoResultTextForFindings"],Self["FNHistoResultTextForFindings"],Self["Findings"],Self["FindingsAlgoNegs"],Self["TP"],Self["FP"],Self["TN"],Self["FN"])
SelfForSheet<-SelfForSheet[!is.na(SelfForSheet$Histo_ResultTextForFindings),]
#Split the dataset up according to the Endoscopist
Dxlst<-split(Self,Self$Endo_Endoscopist)
names(Dxlst)<-paste0('df',gsub("\\s+","",names(Dxlst)))
#Gives the basic output of the number for each indication
numDx<-data.frame(t(lapply(Dxlst,nrow)))
########REFER TO INDIVIDUAL dataframe from the list
#x<-Dxlst[["dfSebastianZeki"]]
################## Basic demographics for indications:
Self$IndicationsFroExamination<-gsub("PROCEDURE PERFORMED","",Self$IndicationsFroExamination)
#Include this next line if you want to include the endoscopies where there are multiple indications. The default is now only to include
#endoscopies where there are single indications so that the data is purified
s<-Self[!grepl("\\..*[A-Z]",Self$IndicationsFroExamination),]
s<-strsplit(as.character(Self[["IndicationsFroExamination"]]),'\\.')
IndicVsBx<-data.frame(Reason=unlist(s),AB=rep(Self[["NumbOfBxs"]],sapply(s,FUN=length)))
IndicVsBx$Reason<-as.character(IndicVsBx$Reason)
IndicVsBx<-IndicVsBx[!grepl("\n$",IndicVsBx$Reason),]
IndicVsBx$Reason<-gsub("\n","",IndicVsBx$Reason)
IndicVsBx<-IndicVsBx[!grepl("^$",IndicVsBx$Reason),]
cc<-IndicVsBx %>%
group_by(Reason) %>%
dplyr::summarise(mean=mean(AB,na.rm=T),count=n(),ToTalNumBx=sum(AB,na.rm=T))%>%
filter(count>1,mean>0,!is.na(Reason))
cc<-cc[!is.na(cc$mean),]
myIndications<-c("Abdominal Pain",
"Abdominal Pain/Bloating",
"Altered bowel habit- alternating/constipation",
"Anaemia/Low Iron or Vitamins",
"Blood PR",
"CHRONIC DIARRHOEA",
"Dyspepsia",
"Dysphagia/Odynophagia",
"Known IBD- Assessment of Activity",
"Nausea and/or Vomiting",
"Reflux-like Symptoms/Atypical Chest Pain",
"Surveillance- Family History of CRC/Polyps",
"Surveillance- Known IBD","Surveillance- Previous Colorectal Cancer",
"Surveillance- Previous Polyps","Surveillance- Previous Polyps/Metaplasia etc",
"Weight Loss")
detach(package:plyr)
source("S:\\Gastroenterology\\Seb\\R\\Scripts\\Generics\\Analytics.R")
########################### Number of biopsies taken by person by Indication
IndicVsBxByPerson<-data.frame(Reason=unlist(s),AB=rep(Self[["Endo_Endoscopist"]],sapply(s,FUN=length)),AC=rep(Self[["NumbOfBxs"]],sapply(s,FUN=length)))
IndicVsBxByPerson$Reason<-as.character(IndicVsBxByPerson$Reason)
IndicVsBxByPerson<-IndicVsBxByPerson[!grepl("\n$",IndicVsBxByPerson$Reason),]
IndicVsBxByPerson$Reason<-gsub("\n","",IndicVsBxByPerson$Reason)
IndicVsBxByPerson<-IndicVsBxByPerson[!grepl("^$",IndicVsBxByPerson$Reason),]
ccByPerson<-IndicVsBxByPerson %>%
group_by(Reason,AB) %>%
dplyr::summarise(mean=mean(AC,na.rm=T),count=n())%>%
filter(count>1,mean>0,!is.na(Reason))%>%
mutate(ToTalNumBx =count*mean)
ccByPerson<-data.frame(ccByPerson)
ccByPerson$AB<-as.character(ccByPerson$AB)
ccByPerson<-ccByPerson[grepl(paste(myIndications, collapse='|'), ccByPerson$Reason,perl=TRUE),]
BxByIndicationByPesron<-dcast(ccByPerson,formula=Reason~AB,value.var="mean")
#Get the mean number of biopsies for each indication for the unit so can compare to individual's number of biopsies in the individual function
row.names(BxByIndicationByPesron)<-BxByIndicationByPesron$Reason
#BxByIndicationByPesron<-BxByIndicationByPesron[-grepl("\n",row.names(BxByIndicationByPesron)),]
BxByIndicationByPesron$Reason<-NULL
MeanNumBx<-data.frame(apply(BxByIndicationByPesron,1,function(x){mean(x,na.rm=T)}))
#row.names(MeanNumBx)<-gsub("\n","",row.names(MeanNumBx))
names(MeanNumBx)<-c("Result")
MeanNumBx$Reason<-row.names(MeanNumBx)
MeanNumBx<-MeanNumBx[grepl(paste(myIndications, collapse='|'), MeanNumBx$Reason,perl=TRUE),]
########################### Number of endoscopies including subtype for this month##############################################################
Self$ProcPerformed<-as.character(Self$ProcPerformed)
ProcedurePlot<-Self %>%
group_by(ProcPerformed) %>%
summarise(freq=n())
#Getting heatmap of endoscopy types done by person
matProcPerf<-as.matrix(table(Self$ProcPerformed,Self$Endo_Endoscopist))
myEndoMatrix<-as.data.frame(table(Self$ProcPerformed,Self$Endo_Endoscopist))
myEndoMatrix<-xtabs(Freq~Var1+Var2,data=myEndoMatrix)
#Order the bar chart
ProcedurePlot$ProcPerformed <- factor(ProcedurePlot$ProcPerformed, levels = ProcedurePlot$ProcPerformed[order(ProcedurePlot$freq)])
#Do the chart for the biopsies using only the ones that are across all the endoscopist (from MeanNumBx)
ccByPerson$Reason<-gsub("\n","",ccByPerson$Reason)
ccByPerson<-data.frame(ccByPerson[which(ccByPerson$Reason %in% MeanNumBx$Reason),])
matPoss<-xtabs(mean~Reason+AB,data=ccByPerson)
############################# Faceted case mix ##########################################################
ProcedurePlot2<-Self %>%
group_by(ProcPerformed,Endo_Endoscopist) %>%
summarise(freq=n())
p=ggplot(data=ProcedurePlot2,aes(x=ProcPerformed,freq,fill=Endo_Endoscopist))+
geom_bar(stat="identity")+
#coord_polar(theta="y")+
theme(legend.position="none")+
theme(axis.text.x=element_text(angle=-90))
rmarkdown::render("S:\\Gastroenterology\\Seb\\R\\Scripts\\SelfAudit\\Self_Unit.Rmd",output_file=paste("MonthlyEndoscopyUnitReport",Sys.Date(),'.docx'))
###################################### Whole Unit path diagnosis rate for use in mydf and final report
DxlstUnit<-as.list(split(Self,Self$Endo_ResultName))
ProcDiagnosisUnit<-function(n,DxlstUnit) {
#Note you pass the names of the dataframes in the list to the function and then you extract the dataframe from the list. Each dataframe should be a procedure
this_is_a_name <- n
this_is_my_data <- DxlstUnit[[n]]
#The corpus to cover both path and endoscopy
ifelse(grepl("Flexible Sigmoidoscopy$",this_is_a_name,perl=TRUE),myNotableWords<-c("[Cc]olitis","[Cc]rohn","[Hh]yperplastic","[Ii]nflammation","[Ii]schaemic","[Mm]icroscopic"),
ifelse(grepl("Gastroscopy \\(OGD\\)$",this_is_a_name,perl=TRUE),myNotableWords<-c("[Bb]arrett","[Cc]andida","[Cc]oeliac","[Ee]osinophilic","[Pp]eptic"),
ifelse(grepl("Colonoscopy$",this_is_a_name,perl=TRUE),myNotableWords<-c("[Cc]olitis","[Cc]rohn","[Hh]yperplastic","[Ii]nflammation","[Ii]schaemic","[Mm]icroscopic"),
myNotableWords<-c("Nothing"))))
if(myNotableWords!="Nothing"){
DirectedDocumentationNoGraph(this_is_a_name,this_is_my_data,"Dx",myNotableWords)
}
}
mydf<-lapply(names(DxlstUnit), ProcDiagnosisUnit,DxlstUnit=DxlstUnit)
names(mydf)<-c("Colonoscopy","ERCP","nil","Flexible Sigmoidoscopy","Gastroscopy (OGD)")
#Need to change this so get outputs for everyone
#x <- "dfMsHarrietWatson"
#This function extracts all the metrics for the report. Firstly it extracts the presenting complaints, the indications vs average number of biopsies taken and the paper clinic otuput
ReportOP<-function(x,Dxlst,mydf){
#This is to get the name of the dataframe
this_is_a_name <- x;
print(this_is_a_name)
#This is to make sure that I get the data from the dataframe as I am passing the names into the function
x <- Dxlst[[x]]
#Number of endoscopies done over time
x<-x %>%
mutate(date=as.Date(format(x[["Endo_ResultEntered"]], "%Y/%m/%d")))
x$ProcPerformed<-as.character(x[["ProcPerformed"]])
counts<-t(table(x[["date"]],x[["ProcPerformed"]]))
counts_df<-melt(counts,var.id=counts)
NumDone<-barplot(counts,xlab="Date",main="Procedures performed in the month",ylab="Number procedures",legend=rownames(counts))
tbl<-table(x[["ProcPerformed"]])
x[["IndicationsFroExamination"]]<-gsub("\\.","\n",x[["IndicationsFroExamination"]])
#BDR<-lapply(names(PClst), BDR_ProcDiagnosis,PClst=PClst,)
Gastroscop<-x[x[["ProcPerformed"]]=="Gastroscopy (OGD)",]
GastroForBDR<-x[grepl("Gastroscopy (OGD)",x[["ProcPerformed"]],fixed=T),]
GastroForBDR<-data.frame(GastroForBDR)
try(BDR<-BDR_DirectedDocumentation(GastroForBDR,"Dx"))
BDR<-as.numeric(BDR)
################################################################################################################################
################################################################################################################################
################################################################################################################################
######################## Adenoma detection rate ##############################################################################
################################################################################################################################
################################################################################################################################
Colons<-x[grep("Colonoscopy.*",x[["ProcPerformed"]]),]
ColonsAdr<-Colons[grep(".*[Aa]denom.*",Colons$Dx),]
ADR<-(nrow(ColonsAdr)/nrow(Colons))*100
################################################################################################################################
################################################################################################################################
################################################################################################################################
######################## Diagnoses made by endoscopy type ####################################################################
################################################################################################################################
################################################################################################################################
#This is how to split into the different endoscopies
PClst<-split(x,x$ProcPerformed)
#names(PClst)<-paste0('df',gsub("\\s+","",names(PClst)))
#mydf
z<-lapply(names(PClst), ProcDiagnosis,PClst=PClst,mydf=mydf)
#Get rid of null lists which mess up grid.arrange
z<-z[lapply(z,length)>0]
################################################################################################################################
################################################################################################################################
################################################################################################################################
#########################Number of biopsies vs indication #####################################################################
################################################################################################################################
################################################################################################################################
x[["IndicationsFroExamination"]]<-gsub("PROCEDURE PERFORMED","",x[["IndicationsFroExamination"]])
numbx<-data.frame(x[["IndicationsFroExamination"]],x[["NumbOfBxs"]])
names(numbx)<-c("IndicationsFroExamination","NumbOfBxs")
grp <- numbx %>% group_by(IndicationsFroExamination) %>% summarise(mean = mean(NumbOfBxs))
########FILTER FOR PURE INDICATIONS
#Only chronic diarrhoea on its own
#Only dysphagia on its own
IndicVsBx<-data.frame(cSplit(x,"IndicationsFroExamination","\n","long")[, .(Avg=mean(NumbOfBxs)),.(IndicationsFroExamination)])
s<-strsplit(as.character(x[["IndicationsFroExamination"]]),'\n')
IndicVsBx<-data.frame(Reason=unlist(s),AB=rep(x[["NumbOfBxs"]],sapply(s,FUN=length)))
IndicVsBx$Reason<-as.character(IndicVsBx$Reason)
#View(IndicVsBx)
IndicVsBx2<-IndicVsBx %>%
group_by(Reason) %>%
dplyr::summarise(mean=mean(AB,na.rm=T))
IndicVsBx2<-IndicVsBx2[!is.na(IndicVsBx2$mean),]
#merge with the relevant fromt eh unit averages from
IndicVsBx2<-IndicVsBx2[!grepl("\n$",IndicVsBx2$Reason),]
IndicVsBx2$Reason<-gsub("\n ","",IndicVsBx2$Reason)
IndicVsBx2$Reason<-gsub("\n","",IndicVsBx2$Reason)
IndicVsBx3<-inner_join(MeanNumBx,IndicVsBx2,by=c("Reason"))
IndicVsBx3<-IndicVsBx3[!is.na(IndicVsBx3$mean),]
IndicVsBx3<-IndicVsBx3[IndicVsBx3$Reason!="",]
################################################################################################################################
################################################################################################################################
################################################################################################################################
####################A##### Separate out the diagnoses and create a table for each one ###########################################
################################################################################################################################
################################################################################################################################
#Build a heatmap of each Indication and diagnosis on biopsy:# Ditch this for now in the final report but maybe use later
#Splits the frame so each indication has its own row
try(dfByIndicPerRow<-data.frame(cSplit(x,"IndicationsFroExamination","\n","long")))
#PClst<-split(dfByIndicPerRow,dfByIndicPerRow$Endo_ResultName)
try(PClst<-split(dfByIndicPerRow,dfByIndicPerRow$IndicationsFroExamination))
try(zz<-lapply(names(PClst), ProcDiagnosis2,PClst=PClst))
try(names(zz)<-names(PClst))
try(zz<-data.frame(Filter(Negate(function(x) is.null(unlist(x))),zz)))
try(rownames(zz)<-zz[,1])
try(zzz<-zz[grep("Prop",colnames(zz))])
try(colnames(zzz)<-gsub("\\.Prop","",colnames(zzz)))
try(colnames(zzz)<-gsub("\\."," ",colnames(zzz)))
try(zzz$Dx<-rownames(zzz))
try(zzzLong<-melt(zzz,id=c("Dx")))
try(myzzzMatrix<-xtabs(value~Dx+variable,data=zzzLong))
try(print(myzzzMatrix))
Cancertbl<-data.frame(x[!is.na(x[["Cancer"]]),])
Cancertbl<-data.frame(Cancertbl["PatientID"],Cancertbl["PatientName"],Cancertbl["Findings"],Cancertbl["DxRaw"])
Dysplasiatbl<-x[!is.na(x[["Dysplasia"]]),]
Dysplasiatbl<-data.frame(Dysplasiatbl["PatientID"],Dysplasiatbl["PatientName"],Dysplasiatbl["Findings"],Dysplasiatbl["DxRaw"])
#######This is where to merge the AllTbl with the image table so you can see the images themselves ####### ####### #######
####### ####### ####### ####### ####### ####### ####### ####### ####### ####### ####### ####### ####### #######
x<-x[!is.na(x$Histol),]
row.names(x) <- NULL
Alltbl<-data.frame(x["PatientID"],x["PatientName"],x["Findings"],x["DxRaw"],x["url"])
#Alltbl<-x
Alltbl$ID<-paste(apply( Alltbl[1:2] , 1 , paste , collapse = " ",sep="\n" ))
Alltbl$ID<-gsub(",",",\n",Alltbl$ID)
Alltbl<-data.frame(Alltbl["ID"],Alltbl["Findings"],Alltbl["DxRaw"],Alltbl["url"])
try(FinalTable2<-GRS_Type_Assess_By_Unit(x))
rmarkdown::render("S:\\Gastroenterology\\Seb\\R\\Scripts\\SelfAudit\\Self.Rmd",params=list(this_is_a_name),output_file=paste(this_is_a_name,"report",Sys.Date(),'.docx'))
print("done")
}
#This gets the unit's figures as a whole
#To get the units pathological diagnosis rate by each type of endoscopy:
lapply(names(Dxlst),ReportOP,Dxlst=Dxlst,mydf=mydf)
#To do:
#Conclusion
sebastiz/SelfAwaRe documentation built on May 29, 2019, 11:04 p.m.
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Defines functions ProcDiagnosisUnit ReportOP
library(stringr)
library(ggplot2)
library(lattice)
library(compare)
library(XLConnect)
library(grid)
library(reshape2)
library(gtools)
library(gplots)
library("tm")
library("SnowballC")
library("wordcloud")
library("RColorBrewer")
library(splitstackshape)
library(knitr)
library(xtable)
library(pander)
library(data.table)
library(gridExtra)
library(pheatmap)
library(tidyr)
library(EndoMineR)
# local({
# hook_plot = knit_hooks$get('plot')
# knit_hooks$set(plot = function(x, options) {
# x = paste(x, collapse = '.')
# if (!grepl('\\.svg', x)) return(hook_plot(x, options))
# # read the content of the svg image and write it out without <?xml ... ?>
# paste(readLines(x)[-1], collapse = '\n')
# #paste("<figure><img src=\"", opts_knit$get("base.url"), paste(x, collapse = "."),
# # "\"><figcaption>", options$fig.cap, "</figcaption></figure>", sep = "")
# })
# })
#For the imag extraction- need to get this off EndoSoft. Choose the query called Images Captured with Proc Data Audit for the month beginning 1 to the end of the month
#Then export HTML to folder called "S:\Gastroenterology\Seb\R\Data\SelfAudit" with styles and Fixed width and JPEG (default settings selected) and it will create a folder called Images Captured with Proc Data Audit.files. The html will be one level up
############################################# ############ DATA ACQUISITION ################## ############################################# #############################################
Self = loadWorkbook("S:\\Gastroenterology\\Seb\\R\\Data\\SelfAudit\\Gastroscopy_TCR2232HistoReport_run%2004-01-2017(Apr18).xlsx")
Self = XLConnect::readWorksheet(Self, sheet="Data",header=TRUE)
############################################# ############ DATA CLEANING ################## ############################################# #############################################
names(Self)<-c("PatientID", "NHSno", "PatientName", "birthdaynum","birthmonthnum" ,"birthyearnum" ,"Endo_ResultName", "Endo_ResultPerformed", "Endo_ResultEntered", "Endo_ResultText", "Histo_ResultName", "Histo_ResultPerformed", "Histo_ResultEntered", "Histo_ResultText")
#Remove the first row if necessary
Self<-Self[2:nrow(Self),]
#Get rid of everything without an endoscopy report (because pre 2007 there won't be one, just histopath)
Self<-subset(Self,!(is.na(Self["Endo_ResultName"])))
#Split the spreadsheet up into new columns and then get rid of the original ones
#For the Endoscopy text
source("S:\\Gastroenterology\\Seb\\R\\Scripts\\Generics\\CleanUp.R")
#Self<-EndoscChopper(Self)
Self$Endo_ResultText<-gsub("2nd Endoscopist:","Second endoscopist:",Self$Endo_ResultText)
EndoscTree<-list("Patient Name:","Date of Birth:","General Practicioner:","Hospital Number:",
"Date of Procedure:","Endoscopist:","Second endoscopist:","Trainee:","Referring Physician:","Nurses:","Medications",
"Instrument","Extent of Exam:","INDICATIONS FOR EXAMINATION","PROCEDURE PERFORMED","FINDINGS",
"ENDOSCOPIC DIAGNOSIS","RECOMMENDATIONS","COMMENTS","FOLLOW UP","")
for(i in 1:(length(EndoscTree)-1)) {
Self<-Extractor(Self,"Endo_ResultText",as.character(EndoscTree[i]),
as.character(EndoscTree[i+1]),as.character(EndoscTree[i]))
}
#Extra tidying...DATES ARE A PAIN IN THE ASS....
Self$PROCEDUREPERFORMED<-gsub("(Withdrawal|Quality|ENDOSCOPIC).*","",Self$PROCEDUREPERFORMED)
Self$PROCEDUREPERFORMED<-gsub("\\s+[(]OGD[)]","",Self$PROCEDUREPERFORMED)
#Convert date time to date in R
Self$Endo_ResultEntered<-as.Date(as.POSIXct(Self$Endo_ResultEntered,format="%d/%m/%Y"))
#And sometimes this is the one to use...
#Self$Endo_ResultEntered<-as.Date(format(Self$Endo_ResultEntered), "%d/%m/%Y")
#Change the names of the columns as EPR keep on changing things around:
names(Self)<-c("PatientID","NHSno","PatientName","birthdaynum","birthmonthnum","birthyearnum","Endo_ResultName","Endo_ResultPerformed","Endo_ResultEntered","Endo_ResultText","Histo_ResultName","Histo_ResultPerformed","Histo_ResultEntered","Histo_ResultText","DateofBirth","GeneralPracticioner","HospitalNumber","DateofProcedure","Endo_Endoscopist","SecondEndo_Endoscopist","Trainee","ReferringPhysician","Nurses","Medications","Instrument","ExtentofExam","IndicationsFroExamination","ProcPerformed","Findings","EndoscopDiagn","Recommend","COMMENTS","FollowUp")
Self<-HistolChopper(Self)
Self$FindingsAlgoNegs<-str_extract_all(Self$Findings,".*[Nn]egative.*|.*[Nn]either.*|[Nn]o .*?[\\.\n]|.*[Nn]ormal.*?[\\.\n]|There (?:is|was|are) no.*?\\.|.*None.*?[\\.\\n]|.*unremarkable?\\.|^Negative.*?[\\.\\n]|[Oo]therwise (?:normal|unremarkable).*?[\\.\\n]")
Self$EndoscopDiagnAlgoNegs<-str_extract_all(Self$EndoscopDiagn,".*[Nn]egative.*|.*[Nn]either.*|[Nn]o .*?[\\.\\n]|.*[Nn]ormal.*?[\\.\\n]|There (?:is|was|are) no.*?\\.|.*None.*?[\\.\\n]|.*unremarkable?\\.|^Negative.*?[\\.\\n]|[Oo]therwise (?:normal|unremarkable).*?[\\.\\n]")
#Get histopath into a proper format for line by line extraction:
Self$Histo_ResultTextForFindings<-gsub("\\.\\.","\\.",Self$Histo_ResultTextForFindings)
Self$Histo_ResultTextForFindings<-gsub("\\.","\\.\n",Self$Histo_ResultTextForFindings)
Self$Histo_ResultTextForFindings<-gsub(":","",Self$Histo_ResultTextForFindings)
Self$Histo_ResultTextForFindingsAlgoNegs<-str_extract_all(Self$Histo_ResultTextForFindings,".*[Nn]egative.*|.*[Nn]either.*|[Nn]o .*?[\\.\\n]|.*[Nn]ormal.*?[\\.\\n]|There (?:is|was|are) no.*?[\\.\\n]|\\..*None.*?[\\.\\n]|\\..*unremarkable.*?[\\.\\n]|^Negative.*?[\\.\\n]|[Oo]therwise (?:normal|unremarkable).*?[\\.\\n]")
SelfWithNegs<-data.frame(Self["PatientID"],Self["Findings"],Self["FindingsAlgoNegs"],Self["EndoscopDiagn"],Self["EndoscopDiagnAlgoNegs"],Self["Histo_ResultTextForFindings"],Self["Histo_ResultTextForFindingsAlgoNegs"])
#Import the image file here from html report from Endosoft:
#You need to go to Endosoft, then pick the query under the GRS tab, under GSTT, called Images Captured with Proc Data Audit.files
#Then run for the month you are interested in
#Then export as HTML into the folder called "S:\Gastroenterology\Seb\R\Data\SelfAudit" It will name the folder itself as Images Captured with Proc Data Audit.files and the html file will be in the SelfAudit level
#Then run the script
############################### Drop in the image table here############################### ###############################
source("S:\\Gastroenterology\\Seb\\R\\Scripts\\SelfAudit\\SelfimgExtractionScripts.R")
images<-MyImgLibrary()
source("S:\\Gastroenterology\\Seb\\R\\Scripts\\Generics\\CleanUp.R")
Self<-merge(images,Self,by=c("Endo_ResultEntered","PatientID"),all.y=T)
############################################# ######## DATA CLEANING ###################### #############################################
#Tidy the whole thing up:
Self$Endo_Endoscopist.x<-NULL
names(Self)[names(Self) == 'Endo_Endoscopist.y'] <-"Endo_Endoscopist"
Self$V1.x<-NULL
Self$V1<-NULL
rmarkdown::render("S:\\Gastroenterology\\Seb\\R\\Scripts\\SelfAudit\\Self2SANDBOX.Rmd",output_file=paste("MonthlyEndoscopyUnitReport2",Sys.Date(),'.docx'))
#Count the number of sentences in a column:
#Get rid of the starting period
Self$Findings<-gsub(".\n.\n",".\n",Self$Findings)
Self$TotalLines<-str_count(Self$Findings, ".\n")
Self$TP<-str_count(Self$FindingsAlgoNegs, "\\.")
Self$TN<-Self$TotalLines-Self$TP
Self$FP<-0
Self$FN<-0
#Self$Histo_ResultTextForFindings2<-gsub("\\.",".\n",Self$Histo_ResultTextForFindings)
Self$TotalLinesHistoResultTextForFindings<-str_count(Self$Histo_ResultTextForFindings, ".\n")
Self$TPHistoResultTextForFindings<-str_count(Self$Histo_ResultTextForFindingsAlgoNegs, "\\.")
Self$TNHistoResultTextForFindings<-Self$TotalLinesHistoResultTextForFindings-Self$TPHistoResultTextForFindings
Self$FPHistoResultTextForFindings<-0
Self$FNHistoResultTextForFindings<-0
SelfForSheet<-data.frame(Self["Histo_ResultTextForFindings"],Self["Histo_ResultTextForFindingsAlgoNegs"],Self["TotalLinesHistoResultTextForFindings"],Self["TPHistoResultTextForFindings"],Self["FPHistoResultTextForFindings"],Self["TNHistoResultTextForFindings"],Self["FNHistoResultTextForFindings"],Self["Findings"],Self["FindingsAlgoNegs"],Self["TP"],Self["FP"],Self["TN"],Self["FN"])
SelfForSheet<-SelfForSheet[!is.na(SelfForSheet$Histo_ResultTextForFindings),]
#Split the dataset up according to the Endoscopist
Dxlst<-split(Self,Self$Endo_Endoscopist)
names(Dxlst)<-paste0('df',gsub("\\s+","",names(Dxlst)))
#Gives the basic output of the number for each indication
numDx<-data.frame(t(lapply(Dxlst,nrow)))
########REFER TO INDIVIDUAL dataframe from the list
#x<-Dxlst[["dfSebastianZeki"]]
################## Basic demographics for indications:
Self$IndicationsFroExamination<-gsub("PROCEDURE PERFORMED","",Self$IndicationsFroExamination)
#Include this next line if you want to include the endoscopies where there are multiple indications. The default is now only to include
#endoscopies where there are single indications so that the data is purified
s<-Self[!grepl("\\..*[A-Z]",Self$IndicationsFroExamination),]
s<-strsplit(as.character(Self[["IndicationsFroExamination"]]),'\\.')
IndicVsBx<-data.frame(Reason=unlist(s),AB=rep(Self[["NumbOfBxs"]],sapply(s,FUN=length)))
IndicVsBx$Reason<-as.character(IndicVsBx$Reason)
IndicVsBx<-IndicVsBx[!grepl("\n$",IndicVsBx$Reason),]
IndicVsBx$Reason<-gsub("\n","",IndicVsBx$Reason)
IndicVsBx<-IndicVsBx[!grepl("^$",IndicVsBx$Reason),]
cc<-IndicVsBx %>%
group_by(Reason) %>%
dplyr::summarise(mean=mean(AB,na.rm=T),count=n(),ToTalNumBx=sum(AB,na.rm=T))%>%
filter(count>1,mean>0,!is.na(Reason))
cc<-cc[!is.na(cc$mean),]
myIndications<-c("Abdominal Pain",
"Abdominal Pain/Bloating",
"Altered bowel habit- alternating/constipation",
"Anaemia/Low Iron or Vitamins",
"Blood PR",
"CHRONIC DIARRHOEA",
"Dyspepsia",
"Dysphagia/Odynophagia",
"Known IBD- Assessment of Activity",
"Nausea and/or Vomiting",
"Reflux-like Symptoms/Atypical Chest Pain",
"Surveillance- Family History of CRC/Polyps",
"Surveillance- Known IBD","Surveillance- Previous Colorectal Cancer",
"Surveillance- Previous Polyps","Surveillance- Previous Polyps/Metaplasia etc",
"Weight Loss")
detach(package:plyr)
source("S:\\Gastroenterology\\Seb\\R\\Scripts\\Generics\\Analytics.R")
########################### Number of biopsies taken by person by Indication
IndicVsBxByPerson<-data.frame(Reason=unlist(s),AB=rep(Self[["Endo_Endoscopist"]],sapply(s,FUN=length)),AC=rep(Self[["NumbOfBxs"]],sapply(s,FUN=length)))
IndicVsBxByPerson$Reason<-as.character(IndicVsBxByPerson$Reason)
IndicVsBxByPerson<-IndicVsBxByPerson[!grepl("\n$",IndicVsBxByPerson$Reason),]
IndicVsBxByPerson$Reason<-gsub("\n","",IndicVsBxByPerson$Reason)
IndicVsBxByPerson<-IndicVsBxByPerson[!grepl("^$",IndicVsBxByPerson$Reason),]
ccByPerson<-IndicVsBxByPerson %>%
group_by(Reason,AB) %>%
dplyr::summarise(mean=mean(AC,na.rm=T),count=n())%>%
filter(count>1,mean>0,!is.na(Reason))%>%
mutate(ToTalNumBx =count*mean)
ccByPerson<-data.frame(ccByPerson)
ccByPerson$AB<-as.character(ccByPerson$AB)
ccByPerson<-ccByPerson[grepl(paste(myIndications, collapse='|'), ccByPerson$Reason,perl=TRUE),]
BxByIndicationByPesron<-dcast(ccByPerson,formula=Reason~AB,value.var="mean")
#Get the mean number of biopsies for each indication for the unit so can compare to individual's number of biopsies in the individual function
row.names(BxByIndicationByPesron)<-BxByIndicationByPesron$Reason
#BxByIndicationByPesron<-BxByIndicationByPesron[-grepl("\n",row.names(BxByIndicationByPesron)),]
BxByIndicationByPesron$Reason<-NULL
MeanNumBx<-data.frame(apply(BxByIndicationByPesron,1,function(x){mean(x,na.rm=T)}))
#row.names(MeanNumBx)<-gsub("\n","",row.names(MeanNumBx))
names(MeanNumBx)<-c("Result")
MeanNumBx$Reason<-row.names(MeanNumBx)
MeanNumBx<-MeanNumBx[grepl(paste(myIndications, collapse='|'), MeanNumBx$Reason,perl=TRUE),]
########################### Number of endoscopies including subtype for this month##############################################################
Self$ProcPerformed<-as.character(Self$ProcPerformed)
ProcedurePlot<-Self %>%
group_by(ProcPerformed) %>%
summarise(freq=n())
#Getting heatmap of endoscopy types done by person
matProcPerf<-as.matrix(table(Self$ProcPerformed,Self$Endo_Endoscopist))
myEndoMatrix<-as.data.frame(table(Self$ProcPerformed,Self$Endo_Endoscopist))
myEndoMatrix<-xtabs(Freq~Var1+Var2,data=myEndoMatrix)
#Order the bar chart
ProcedurePlot$ProcPerformed <- factor(ProcedurePlot$ProcPerformed, levels = ProcedurePlot$ProcPerformed[order(ProcedurePlot$freq)])
#Do the chart for the biopsies using only the ones that are across all the endoscopist (from MeanNumBx)
ccByPerson$Reason<-gsub("\n","",ccByPerson$Reason)
ccByPerson<-data.frame(ccByPerson[which(ccByPerson$Reason %in% MeanNumBx$Reason),])
matPoss<-xtabs(mean~Reason+AB,data=ccByPerson)
############################# Faceted case mix ##########################################################
ProcedurePlot2<-Self %>%
group_by(ProcPerformed,Endo_Endoscopist) %>%
summarise(freq=n())
p=ggplot(data=ProcedurePlot2,aes(x=ProcPerformed,freq,fill=Endo_Endoscopist))+
geom_bar(stat="identity")+
#coord_polar(theta="y")+
theme(legend.position="none")+
theme(axis.text.x=element_text(angle=-90))
rmarkdown::render("S:\\Gastroenterology\\Seb\\R\\Scripts\\SelfAudit\\Self_Unit.Rmd",output_file=paste("MonthlyEndoscopyUnitReport",Sys.Date(),'.docx'))
###################################### Whole Unit path diagnosis rate for use in mydf and final report
DxlstUnit<-as.list(split(Self,Self$Endo_ResultName))
ProcDiagnosisUnit<-function(n,DxlstUnit) {
#Note you pass the names of the dataframes in the list to the function and then you extract the dataframe from the list. Each dataframe should be a procedure
this_is_a_name <- n
this_is_my_data <- DxlstUnit[[n]]
#The corpus to cover both path and endoscopy
ifelse(grepl("Flexible Sigmoidoscopy$",this_is_a_name,perl=TRUE),myNotableWords<-c("[Cc]olitis","[Cc]rohn","[Hh]yperplastic","[Ii]nflammation","[Ii]schaemic","[Mm]icroscopic"),
ifelse(grepl("Gastroscopy \\(OGD\\)$",this_is_a_name,perl=TRUE),myNotableWords<-c("[Bb]arrett","[Cc]andida","[Cc]oeliac","[Ee]osinophilic","[Pp]eptic"),
ifelse(grepl("Colonoscopy$",this_is_a_name,perl=TRUE),myNotableWords<-c("[Cc]olitis","[Cc]rohn","[Hh]yperplastic","[Ii]nflammation","[Ii]schaemic","[Mm]icroscopic"),
myNotableWords<-c("Nothing"))))
if(myNotableWords!="Nothing"){
DirectedDocumentationNoGraph(this_is_a_name,this_is_my_data,"Dx",myNotableWords)
}
}
mydf<-lapply(names(DxlstUnit), ProcDiagnosisUnit,DxlstUnit=DxlstUnit)
names(mydf)<-c("Colonoscopy","ERCP","nil","Flexible Sigmoidoscopy","Gastroscopy (OGD)")
#Need to change this so get outputs for everyone
#x <- "dfMsHarrietWatson"
#This function extracts all the metrics for the report. Firstly it extracts the presenting complaints, the indications vs average number of biopsies taken and the paper clinic otuput
ReportOP<-function(x,Dxlst,mydf){
#This is to get the name of the dataframe
this_is_a_name <- x;
print(this_is_a_name)
#This is to make sure that I get the data from the dataframe as I am passing the names into the function
x <- Dxlst[[x]]
#Number of endoscopies done over time
x<-x %>%
mutate(date=as.Date(format(x[["Endo_ResultEntered"]], "%Y/%m/%d")))
x$ProcPerformed<-as.character(x[["ProcPerformed"]])
counts<-t(table(x[["date"]],x[["ProcPerformed"]]))
counts_df<-melt(counts,var.id=counts)
NumDone<-barplot(counts,xlab="Date",main="Procedures performed in the month",ylab="Number procedures",legend=rownames(counts))
tbl<-table(x[["ProcPerformed"]])
x[["IndicationsFroExamination"]]<-gsub("\\.","\n",x[["IndicationsFroExamination"]])
#BDR<-lapply(names(PClst), BDR_ProcDiagnosis,PClst=PClst,)
Gastroscop<-x[x[["ProcPerformed"]]=="Gastroscopy (OGD)",]
GastroForBDR<-x[grepl("Gastroscopy (OGD)",x[["ProcPerformed"]],fixed=T),]
GastroForBDR<-data.frame(GastroForBDR)
try(BDR<-BDR_DirectedDocumentation(GastroForBDR,"Dx"))
BDR<-as.numeric(BDR)
################################################################################################################################
################################################################################################################################
################################################################################################################################
######################## Adenoma detection rate ##############################################################################
################################################################################################################################
################################################################################################################################
Colons<-x[grep("Colonoscopy.*",x[["ProcPerformed"]]),]
ColonsAdr<-Colons[grep(".*[Aa]denom.*",Colons$Dx),]
ADR<-(nrow(ColonsAdr)/nrow(Colons))*100
################################################################################################################################
################################################################################################################################
################################################################################################################################
######################## Diagnoses made by endoscopy type ####################################################################
################################################################################################################################
################################################################################################################################
#This is how to split into the different endoscopies
PClst<-split(x,x$ProcPerformed)
#names(PClst)<-paste0('df',gsub("\\s+","",names(PClst)))
#mydf
z<-lapply(names(PClst), ProcDiagnosis,PClst=PClst,mydf=mydf)
#Get rid of null lists which mess up grid.arrange
z<-z[lapply(z,length)>0]
################################################################################################################################
################################################################################################################################
################################################################################################################################
#########################Number of biopsies vs indication #####################################################################
################################################################################################################################
################################################################################################################################
x[["IndicationsFroExamination"]]<-gsub("PROCEDURE PERFORMED","",x[["IndicationsFroExamination"]])
numbx<-data.frame(x[["IndicationsFroExamination"]],x[["NumbOfBxs"]])
names(numbx)<-c("IndicationsFroExamination","NumbOfBxs")
grp <- numbx %>% group_by(IndicationsFroExamination) %>% summarise(mean = mean(NumbOfBxs))
########FILTER FOR PURE INDICATIONS
#Only chronic diarrhoea on its own
#Only dysphagia on its own
IndicVsBx<-data.frame(cSplit(x,"IndicationsFroExamination","\n","long")[, .(Avg=mean(NumbOfBxs)),.(IndicationsFroExamination)])
s<-strsplit(as.character(x[["IndicationsFroExamination"]]),'\n')
IndicVsBx<-data.frame(Reason=unlist(s),AB=rep(x[["NumbOfBxs"]],sapply(s,FUN=length)))
IndicVsBx$Reason<-as.character(IndicVsBx$Reason)
#View(IndicVsBx)
IndicVsBx2<-IndicVsBx %>%
group_by(Reason) %>%
dplyr::summarise(mean=mean(AB,na.rm=T))
IndicVsBx2<-IndicVsBx2[!is.na(IndicVsBx2$mean),]
#merge with the relevant fromt eh unit averages from
IndicVsBx2<-IndicVsBx2[!grepl("\n$",IndicVsBx2$Reason),]
IndicVsBx2$Reason<-gsub("\n ","",IndicVsBx2$Reason)
IndicVsBx2$Reason<-gsub("\n","",IndicVsBx2$Reason)
IndicVsBx3<-inner_join(MeanNumBx,IndicVsBx2,by=c("Reason"))
IndicVsBx3<-IndicVsBx3[!is.na(IndicVsBx3$mean),]
IndicVsBx3<-IndicVsBx3[IndicVsBx3$Reason!="",]
################################################################################################################################
################################################################################################################################
################################################################################################################################
####################A##### Separate out the diagnoses and create a table for each one ###########################################
################################################################################################################################
################################################################################################################################
#Build a heatmap of each Indication and diagnosis on biopsy:# Ditch this for now in the final report but maybe use later
#Splits the frame so each indication has its own row
try(dfByIndicPerRow<-data.frame(cSplit(x,"IndicationsFroExamination","\n","long")))
#PClst<-split(dfByIndicPerRow,dfByIndicPerRow$Endo_ResultName)
try(PClst<-split(dfByIndicPerRow,dfByIndicPerRow$IndicationsFroExamination))
try(zz<-lapply(names(PClst), ProcDiagnosis2,PClst=PClst))
try(names(zz)<-names(PClst))
try(zz<-data.frame(Filter(Negate(function(x) is.null(unlist(x))),zz)))
try(rownames(zz)<-zz[,1])
try(zzz<-zz[grep("Prop",colnames(zz))])
try(colnames(zzz)<-gsub("\\.Prop","",colnames(zzz)))
try(colnames(zzz)<-gsub("\\."," ",colnames(zzz)))
try(zzz$Dx<-rownames(zzz))
try(zzzLong<-melt(zzz,id=c("Dx")))
try(myzzzMatrix<-xtabs(value~Dx+variable,data=zzzLong))
try(print(myzzzMatrix))
Cancertbl<-data.frame(x[!is.na(x[["Cancer"]]),])
Cancertbl<-data.frame(Cancertbl["PatientID"],Cancertbl["PatientName"],Cancertbl["Findings"],Cancertbl["DxRaw"])
Dysplasiatbl<-x[!is.na(x[["Dysplasia"]]),]
Dysplasiatbl<-data.frame(Dysplasiatbl["PatientID"],Dysplasiatbl["PatientName"],Dysplasiatbl["Findings"],Dysplasiatbl["DxRaw"])
#######This is where to merge the AllTbl with the image table so you can see the images themselves ####### ####### #######
####### ####### ####### ####### ####### ####### ####### ####### ####### ####### ####### ####### ####### #######
x<-x[!is.na(x$Histol),]
row.names(x) <- NULL
Alltbl<-data.frame(x["PatientID"],x["PatientName"],x["Findings"],x["DxRaw"],x["url"])
#Alltbl<-x
Alltbl$ID<-paste(apply( Alltbl[1:2] , 1 , paste , collapse = " ",sep="\n" ))
Alltbl$ID<-gsub(",",",\n",Alltbl$ID)
Alltbl<-data.frame(Alltbl["ID"],Alltbl["Findings"],Alltbl["DxRaw"],Alltbl["url"])
try(FinalTable2<-GRS_Type_Assess_By_Unit(x))
rmarkdown::render("S:\\Gastroenterology\\Seb\\R\\Scripts\\SelfAudit\\Self.Rmd",params=list(this_is_a_name),output_file=paste(this_is_a_name,"report",Sys.Date(),'.docx'))
print("done")
}
#This gets the unit's figures as a whole
#To get the units pathological diagnosis rate by each type of endoscopy:
lapply(names(Dxlst),ReportOP,Dxlst=Dxlst,mydf=mydf)
#To do:
#Conclusion
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Embedding an R snippet on your website
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.