lilo

################################### ####Writing data into excel function######### ################################### #Method one # Write the first data set in a new workbook write.xlsx(Data1, file=”exportedata.xlsx”,sheetName=”USA-ARRESTS”, append=FALSE) # Add a second data set in a new worksheet write.xlsx(Data2, file=”exportedata.xlsx”, sheetName=”MTCARS”,append=TRUE) # Add a third data set write.xlsx(Data3, file=”exportedata.xlsx”, sheetName=”TITANIC”,append=TRUE)   #Method two # file : the path to the output file … Read more

##Association rule or Market basket library(“arules”) removewords=c(names(termFrequency)[which(termFrequency==1)],’en’,’f’,’nicht’,’es’,’luck’,’giving’,’thought’,’value’,’indeed’,’almost’,’apparently’,’exist’,’d’,’net’,’ture’,’dans’,’des’,’et’,’ne’,’une’,’le’) VerbList=sapply(DATA_.input, function(x){strsplit(x[[1]],’ ‘)}) VerbList=sapply(VerbList, function(x){ Idx=which(x==”” | x %in% removewords) if(length(Idx)>0)x=x[-Idx] else x=x x=unique(x)} ) VerbList=sapply(VerbList,function(x){ paste(x,collapse=’,’)}) temp=which(VerbList==”) VerbList=VerbList[-temp] head(VerbList) write(VerbList,file=’C:\\Users\\folder\\Desktop\\VerbList_a’) verbWordList<- read.transactions(“C:\\Users\\folder\\Desktop\\VerbList_a”, format=”basket”,sep=”,”) rules <- apriori(verbWordList, parameter = list(support = 0.01,confidence = 0.01,minlen=2)) rules.sorted <- sort(rules, by=”support”) inspect(rules.sorted) #inspect(rules.sorted[1:5]) if(length(rules.sorted)>0){ rules.table=list(Keywords=lapply(1:length(rules.sorted), function(i){ wlist=do.call(‘c’,c(LIST(lhs(rules.sorted[i])),LIST(rhs(rules.sorted[i])))) }), quality=quality(rules.sorted)) } #changed … Read more

# word cloud is easy #we need cleanDescription.r file as follows. It was used to write some similar words into one word. For example, write games and gaming into game. You can write lots of them into it. require(tm) cleanDescription <- function(description,additional.stopwords=NULL) { # write to lower case description <- tolower(description) ##remove non-character symbols description … Read more

# we have a data set with all numeric variables (some of the columns only have 0 or 1). And we want to run a logistic regression model. First we checked correlation and exclude the variables which has high correlation (remove one and keep the other). COR<-cor(Answered_data2) data<-subset(data,select=-c(A,C,D,AA,CC)) #we need to run correlation multiple times … Read more

#If some code does not work, we need to figure out if it is because some library is not loaded. Like filter, we need dplyr library to be loaded first. library(dplyr) T<-filter(P_HMD_TEST_PRED,Tabc==1) # we usually using conditions like this A<-filter(data, EE==1 & EE==2), how about we do not want the logic of and, we want … Read more

# R is simple to do data analyst work. If you want to write complicated loop and agrithem, that would be another story. Here is the code using R in a simple way.   #sometimes we need to install new packages, our R studio is running on the server. So it has a path for … Read more

#This is the most commonly used R code at work everyday. #read data from data.txt file and data2.txt file PP_C<-read.delim(“data1.txt”,header=T,sep=’\t’,comment.char=””,quote=””,row.names=NULL, stringsAsFactors= FALSE ) CES<-read.delim(“data2.txt”,header=T,sep=’\t’,comment.char=””,quote=””,row.names=NULL, stringsAsFactors= FALSE ) #check how many rows there nrow(PP_C) nrow(CES) #check how many unique COMM_ID there length(unique(PP_C$COMM_ID)) length(unique(CES$COMM_ID)) #check column names for each file colnames(PP_C) colnames(CES) #select specific columns to be … Read more