--- title: "Support" author: "Scary Scarecrow" date: "1/12/2022" output: html_document --- ```{r setup, include=FALSE} knitr::opts_chunk$set(echo = TRUE) library(readxl) library(dplyr) library(lubridate) library(DT) library(tidyr) mutlstxlrdr<-function(){ for( i in seq_along(sheet.na)){ colnames<-unique(saptemplate[saptemplate$`Sheet Name`==snames[i],]$Header) df<-read.table("", col.names = colnames) assign(snames[i], df) } } do.call(file.remove, list(list.files("./support/errors/mandatory/", full.names = TRUE))) do.call(file.remove, list(list.files("./support/errors/codelist/", full.names = TRUE))) do.call(file.remove, list(list.files("./support/errors/length/", full.names = TRUE))) do.call(file.remove, list(list.files("./support/summary/", full.names = TRUE))) do.call(file.remove, list(list.files("./support/output/", full.names = TRUE))) ``` ## Data transformation workflow Following is the proposed preliminary workflow for the data transformation project. >All file of a segment (support/accounts etc..) should be inside the relevant folder. Each folder should have one folder for all codelist files. All legacy data (one file for each country) should be inside the raw-data folder, named after each country. Another file having field definitions including name of the matching column from the legacy file should also be there. >*Make sure that there are no hidden files inside the directory.* ### Code Lists ```{r Create List of Files, echo=TRUE, message=FALSE, warning=FALSE} filenames <- list.files("./support/CodeList", pattern="*.xlsx", full.names = T) # We can avoid creating a separate directory for code list. But organizing may be difficult. However, this can be explored further if we want transform all the data in one go i.e. not by functions (support, accounts etc.). # File paths print(filenames) ``` Check manually if the above list includes all the codelist files If correct, then read the files. ```{r codelistreader, echo=TRUE, message=FALSE, warning=FALSE} sheet_names<-lapply(filenames, excel_sheets) # Creates a list of the sheet names codelist_files<-NULL for(i in seq_along(filenames)){ a<-lapply(excel_sheets(filenames[[i]]), read_excel, path = filenames[[i]], col_types = "text") # Reads the sheets of the excel files names(a)<-c(sheet_names[[i]]) # Renames them according to the sheet names extracted above codelist_files<-c(codelist_files,a) } # Names of the files imported names(codelist_files) #codelist_files<-unique(codelist_files) codelist_files$Academic_Title ``` ### Templates Let us now extract the data. Below we are reading only one file having all data related to `support` from the legacy system. ```{r readlegacyfilepath, echo=TRUE, message=FALSE, warning=FALSE} oldfilepath<-list.files("./support/raw-data", pattern="*.xls", full.names = T) # Change the path, check pattern print(oldfilepath) ``` Check it the list matches the actual files, manually. ```{r readlegacyfiles, echo=TRUE} old_files<-NULL #read_excel(path = oldfilepath[[i]], sheet = 1) for(i in seq_along(oldfilepath)){ old_files[[i]]<-read_excel(path = oldfilepath[[i]], sheet = 1) } names(old_files)<-gsub("./support/raw-data/","",oldfilepath) ``` *Some errors in the legacy file noticed. Columns with similar or same name exists.* ```{r readSAPtemplate, echo=TRUE, message=FALSE, warning=FALSE} saptemplate<-read_excel("./support/template.xlsx", sheet = "Field_Definitions") # First few rows of the imported data head(saptemplate) ``` *Please note that the format of the tables (sheet) has been slightly changed. Earlier the corresponding sheet name was mentioned in a row before the actual table. Now, all the rows mention the corresponding sheet name. This was done manually for convenience of data extraction* ```{r createmptySAPfiles, echo=TRUE, message=FALSE, warning=FALSE} #orilo<-"en_US.UTF-8" #Sys.setlocale(locale="en_US.UTF-8") strt<-Sys.time() snames <- unique(saptemplate$`Sheet Name`) for (h in seq_along(old_files)) { # Copy original data old.copy <- old_files[[h]] print(paste0(names(old_files[h])," imported")) err.summ<-data.frame(Country=NULL, Name=NULL, Expected=NULL, Actual=NULL) #Error Cal # Creates data frame for each sheet in snames for (i in seq_along(snames)) { print(paste0("Processing ..",snames[i])) # Select the column names from the field description sheet print("Creating template") sel.template.desc <- saptemplate[saptemplate$`Sheet Name` == snames[i], ] print("Creating column names") sel.template.desc.colnames <- sel.template.desc$Header # Create a list by adding values from corresponding legacy data temp <- NULL print("adding values to template ") for (j in seq_along(sel.template.desc.colnames)) { temp[j] <- ifelse(!is.na(sel.template.desc$default[j]), as.character(as.vector(sel.template.desc$default[j])), ifelse( sel.template.desc$oldkey[j]=="NA" | is.na(sel.template.desc$oldkey[j]), NA, as.vector(old.copy[, sel.template.desc$oldkey[j]]) ) ) } # Rename the columns according to field description print("renaming template ") names(temp) <- sel.template.desc.colnames # Create data frame from the list df <- as.data.frame(temp) print("Converted to data frame") # Error summary file Expected<-nrow(df) #Select essential rows print("Identifying essential rows") sel.template.desc |> filter(Mandatory == "Yes") |> pull(Header) -> essential.columns error.mandatory <- NULL error.df<-data.frame(Country=NULL, Name=NULL, Rows=NULL, Expected=NULL) # Operate on essential columns including creation of error file for (k in seq_along(essential.columns)) { if(essential.columns[k]=="International_Version"){ print("Found International Version. Adding 0.") #stop() df$International_Version<-"0" } print("Creating and writing data with missing mandatory values") assign( paste0( "error_mandatory_", substr(names(old_files[h]), 1, 2), "_", snames[i], "_", essential.columns[k] ), df[is.na(df[, essential.columns[k]]), ] ) # TO be saved in error files if(nrow(df[is.na(df[, essential.columns[k]]), ])>0){ write.csv( df[is.na(df[, essential.columns[k]]), ], paste0( "./support/errors/mandatory/", substr(names(old_files[h]), 1, 2), "_", snames[i], "_", essential.columns[k], "_error_mandatory.csv" ), row.names = F, na="" ) } # Error summary file Country<-substr(names(old_files[h]), 1, 2) Name<-snames[i] err.type<-paste0("Missing ",essential.columns[k]) err.count<-nrow(df[is.na(df[, essential.columns[k]]), ]) print("Removing rows with empty essetial columns") df <- df[!is.na(df[, essential.columns[k]]), ] if(err.count>0){ error.df<-rbind(error.df,data.frame(Country=Country, Name=Name, err.type=err.type, err.count=err.count)) #Error cal } } print("Identifying columns associated with codelists") # List of columns that have a codelist codelistcols <- sel.template.desc |> filter(!is.na(`CodeList File Path`)) |> pull(Header) for (k in seq_along(codelistcols)) { if(codelistcols[k]=="International_Version"){ print("Found International Version. Adding 0.") df$International_Version<-"0" } print(paste0("Identifying errors ",codelistcols[k])) def.rows <- which(!df[, codelistcols[k]] %in% c(pull(codelist_files[codelistcols[k]][[1]], Description), NA)) def.n<- df[def.rows, 1] def.rows.val <- df[!df[, codelistcols[k]] %in% c(pull(codelist_files[codelistcols[k]][[1]], Description), NA), codelistcols[k]] def <- data.frame(def.rows, def.n,def.rows.val) if(nrow(def>0)){ assign(paste0( "error_codematch_", substr(names(old_files[1]), 1, 2), "_", snames[i], "_", codelistcols[k] ), def) # TO be saved write.csv( def, paste0( "./support/errors/codelist/", substr(names(old_files[h]), 1, 2), "_", snames[i], "_", codelistcols[k], "_error_codematch_.csv" ), row.names = F, na="" ) } err.type<-paste0("Codelist Mismatch ", codelistcols[k]) #Error cal err.count<-nrow(def) #Error cal if(err.count>0){ error.df<-rbind(error.df,data.frame(Country=Country, Name=Name, err.type=err.type, err.count=err.count)) #Error cal } print(paste0("Removing errors ",codelistcols[k])) # Removes any mismatch df[!df[, codelistcols[k]] %in% c(pull(codelist_files[codelistcols[k]][[1]], Description), NA), codelistcols[k]] <- NA # Matches each column with the corresponding code list and returns the value df[, codelistcols[k]] <- pull(codelist_files[codelistcols[k]][[1]], 2)[match(pull(df, codelistcols[k]), pull(codelist_files[codelistcols[k]][[1]], Description))] } max.length <- as.numeric(sel.template.desc$`Max Length`) dtype <- sel.template.desc$`Data Type` rowval <- NULL ival <- NULL rval <- NULL lenght.issue.df <- NULL # Changing the data class for (k in 1:ncol(df)) { if (dtype[k] == "String") { df[, k] <- as.character(pull(df, k)) } if (dtype[k] == "Boolean") { df[, k] <- as.logical(pull(df, k)) } if (dtype[k] == "DateTime") { df[, k] <- lubridate::ymd_hms(pull(df, k)) } if (dtype[k] == "Time") { df[, k] <- lubridate::hms(pull(df, k)) } # This list will increase and also change based on input date and time formats } print("Rectifying streetname") # Street and House Number if (any(colnames(df) == "Street")) { print("found steet") # stop() df$Streetname<-NA df$HouseNumber<-NA #df |> extract("Street", "(\\D+)(\\d.*)") df<-tidyr::extract(df, "Street", c("Streetname", "HouseNumber"), "(\\D+)(\\d.*)") df <- df |> select(-c("House_Number")) |> rename(Street = Streetname, House_Number = HouseNumber) |> select(all_of(sel.template.desc.colnames)) } # Length Rectification colclasses <- lapply(df, class) print("Rectifying Length") for (k in 1:ncol(df)) { if (colclasses[[k]] == "character") { print("found character column ") rowval <- pull(df, 1) ival <- ifelse(nchar(pull(df, k))== 0 | is.na(nchar(pull(df, k))),1,nchar(pull(df, k))) rval <- max.length[k] # rectifying data length df[, k] <- ifelse(nchar(pull(df, k)) > max.length[k], substring(pull(df, k), 1, max.length[k]), pull(df, k)) } lenght.issue.df <- rbind(lenght.issue.df, data.frame(rowval, ival, rval)) err.type<- paste0("Length error ", colnames(df)[k]) # Error cal err.count<- sum(ival>rval, na.rm = T) # Error cal if(err.count>0){ error.df<-rbind(error.df,data.frame(Country=Country, Name=Name, err.type=err.type, err.count=err.count)) #Error cal } } lenght.issue.df <- dplyr::filter(lenght.issue.df,ival>rval) if(nrow(lenght.issue.df)>0){ write.csv(lenght.issue.df, paste0( "./support/errors/length/", substr(names(old_files[h]), 1, 2), "_", snames[i], "_length_error.csv" ), row.names = F, na="") } assign(snames[i], df) write.csv(df,paste0("./support/output/", substr(names(old_files[h]), 1, 2), "_", snames[i],".csv"), row.names = F, na="", quote=FALSE) if(nrow(error.df)>0){ write.csv(error.df, paste0("./support/summary/",substr(names(old_files[h]), 1, 2), "_", snames[i],"_error",".csv"), row.names = F, na="") # Error write } err.summ<-rbind(err.summ,data.frame(Country=Country, Name=Name, Expected=Expected, Actual=nrow(df))) #Error Cal } write.csv(err.summ, paste0("./support/summary/" ,substr(names(old_files[h]), 1, 2), "_", snames[i],"_sumerror",".csv"), row.names = F, na="") # Error Write } end<-Sys.time() end-strt ``` *The code failed because Department Column appears several times in the data and while importing R renamed them to Department..xx).* *Manually verify if these are the required templates*