4 files changed, 441 insertions, 0 deletions

diff --git a/admin/survey/R/script/createCrosstabulation2.R b/admin/survey/R/script/createCrosstabulation2.R
new file mode 100644
index 0000000..c8562e5
--- /dev/null
+++ b/admin/survey/R/script/createCrosstabulation2.R
@@ -0,0 +1,110 @@
+# CROSSTABULACIJA 2 SPREMENLJIVK

+

+# preberemo parametre 1->folderpath

+params <- commandArgs(trailingOnly=TRUE)

+

+path <- params[1]

+

+# preberemo sfiltrirano tabelo s podatki

+dataTable <- read.table(

+	file = paste(path, 'admin/survey/R/TempData/crosstab_data.tmp', sep=''),

+	sep = ',', 

+	colClasses = 'numeric',

+	comment.char = '',

+	quote = '',

+	header = FALSE

+)

+

+	

+# KROSTABULACIJA

+# iz tabele pobrisemo vrstice z vrednostmi manjsimi od 0 in sfiltriramo vse stolpce ki jih ne rabimo

+dataTableClean <- dataTable[apply(dataTable[,1:2], MARGIN = 1, function(x) all(x > -1)), ]

+

+# izvedemo krostabulacijo 

+crosstabTable <- xtabs(~dataTableClean[,1]+dataTableClean[,2], data=dataTableClean)

+

+# Vrednosti katerim pripadajo frekvence

+vars <- dimnames(crosstabTable)

+

+# izracunamo vsoto

+sumsVrstica <- apply(crosstabTable, c(1), sum)

+sumsStolpec <- apply(crosstabTable, c(2), sum)

+sums <- sum(crosstabTable)

+

+# x^2

+x2 <- summary(crosstabTable);

+x2 <- x2[3]

+

+

+delezCol <- 3

+

+# POVPRECJE

+if(params[2] == '1'){

+	

+	#dataTableAvg <- dataTableClean[sapply(dataTableClean[,3], function(x) all(x > -1)), ]

+	dataTableAvg <- dataTableClean

+	dataTableAvg[,3][dataTableAvg[,3] < 0] <- 0

+

+	avgTable <- xtabs(dataTableAvg[,3]~dataTableAvg[,1]+dataTableAvg[,2], dataTableAvg) / crosstabTable

+	

+	delezCol <- delezCol + 1

+}

+

+

+# DELEZ

+if(length(params) > 2){

+

+	# array, ki vsebuje vrednosti stolpca, za katere racunamo delez

+	delez <- params[3]

+	delez <- unlist(strsplit(delez, ","))

+	delez <- sapply(delez, strtoi)

+	

+	# ce je -1 gre za checxbox stolpce ki imajo samo vrednost 1 in jih je vec

+	if(delez[1] == -1){

+		dataTableDelez <- dataTableClean

+		dataTableDelez[,delezCol][!(rowSums(dataTableDelez[delezCol:ncol(dataTableDelez)]) == ncol(dataTableDelez)-delezCol+1)] <- 0

+		dataTableDelez[,delezCol][(rowSums(dataTableDelez[delezCol:ncol(dataTableDelez)]) == ncol(dataTableDelez)-delezCol+1)] <- 1

+	}

+	else{

+		#dataTableDelez <- dataTableClean[sapply(dataTableClean[,delezCol], function(x) all(x %in% delez)), ]

+		dataTableDelez <- dataTableClean

+		dataTableDelez[,delezCol][!(dataTableDelez[,delezCol] %in% delez)] <- 0

+		dataTableDelez[,delezCol][(dataTableDelez[,delezCol] %in% delez)] <- 1

+	}

+

+	delezTable <- xtabs(dataTableDelez[,delezCol]~dataTableDelez[,1]+dataTableDelez[,2], dataTableDelez) / crosstabTable

+}

+

+

+# vsako variablo (vrstico) pretvorimo v csv string

+vars <- sapply(vars, paste, collapse=",")

+

+# naredimo 2-d tabelo podatkov

+crosstabTable <- ftable(crosstabTable)

+crosstabTable <- sapply(crosstabTable, paste, collapse=", ")

+

+

+# podatke vrnemo v obliki var11,var12,var13..._var2...--freq1_freq2...

+cat(paste(vars, sep="", collapse="_"))

+cat("--")

+cat(paste(crosstabTable, sep="", collapse="_"))

+cat("--")

+cat(paste(sumsVrstica, sep="", collapse="_"))

+cat("--")

+cat(paste(sumsStolpec, sep="", collapse="_"))

+cat("--")

+cat(paste(sums, sep="", collapse="_"))

+cat("--")

+cat(paste(x2, sep="", collapse="_"))

+

+# rezultati povprecja

+if(params[2] == '1'){

+	cat("--")

+	cat(paste(avgTable, sep="", collapse="_"))

+}

+

+# rezultati deleza

+if(length(params) > 2){

+	cat("--")

+	cat(paste(delezTable, sep="", collapse="_"))

+}
\ No newline at end of file
diff --git a/admin/survey/R/script/createCrosstabulation3.R b/admin/survey/R/script/createCrosstabulation3.R
new file mode 100644
index 0000000..13985ae
--- /dev/null
+++ b/admin/survey/R/script/createCrosstabulation3.R
@@ -0,0 +1,108 @@
+# CROSSTABULACIJA 3 SPREMENLJIVK

+

+# preberemo parametre 1->folderpath

+params <- commandArgs(trailingOnly=TRUE)

+

+path <- params[1]

+

+# preberemo sfiltrirano tabelo s podatki

+dataTable <- read.table(

+	file = paste(path, 'admin/survey/R/TempData/crosstab_data.tmp', sep=''),

+	sep = ',', 

+	colClasses = 'numeric',

+	comment.char = '',

+	quote = '',

+	header = FALSE

+)

+

+

+# iz tabele pobrisemo vrstice z vrednostmi manjsimi od 0 in sfiltriramo vse stolpce ki jih ne rabimo

+dataTableClean <- dataTable[apply(dataTable[,1:3], MARGIN = 1, function(x) all(x > -1)), 1:3]

+

+

+

+# izvedemo krostabulacijo

+crosstabTable <- xtabs(~dataTableClean[,1]+dataTableClean[,2]+dataTableClean[,3], data=dataTableClean)

+

+

+

+# Vrednosti katerim pripadajo frekvence

+vars <- dimnames(crosstabTable)

+

+# izracunamo vsoto

+sumsVrstica <- apply(crosstabTable, c(2), sum)

+sumsStolpec <- apply(crosstabTable, c(3), sum)

+sums <- sum(crosstabTable)

+	

+

+delezCol <- 4

+

+# POVPRECJE

+if(params[2] == '1'){

+	

+	#dataTableAvg <- dataTableClean[sapply(dataTableClean[,4], function(x) all(x > -1)), ]

+	dataTableAvg <- dataTableClean

+	dataTableAvg[,4][dataTableAvg[,4] < 0] <- 0

+	

+	avgTable <- xtabs(dataTableAvg[,4]~dataTableAvg[,1]+dataTableAvg[,2]+dataTableAvg[,3], dataTableAvg) / crosstabTable

+	

+	delezCol <- delezCol + 1

+}

+

+

+# DELEZ

+if(length(params) > 2){

+

+	# array, ki vsebuje vrednosti stolpca, za katere racunamo delez

+	delez <- params[3]

+	delez <- unlist(strsplit(delez, ","))

+	delez <- sapply(delez, strtoi)

+	

+	# ce je -1 gre za checxbox stolpce ki imajo samo vrednost 1 in jih je vec

+	if(delez[1] == -1){

+		dataTableDelez <- dataTableClean

+		dataTableDelez[,delezCol][!(rowSums(dataTableDelez[delezCol:ncol(dataTableDelez)]) == ncol(dataTableDelez)-delezCol+1)] <- 0

+		dataTableDelez[,delezCol][(rowSums(dataTableDelez[delezCol:ncol(dataTableDelez)]) == ncol(dataTableDelez)-delezCol+1)] <- 1

+	}

+	else{	

+		#dataTableDelez <- dataTableClean[sapply(dataTableClean[,delezCol], function(x) all(x %in% delez)), ]

+		dataTableDelez[,delezCol][!(dataTableDelez[,delezCol] %in% delez)] <- 0

+		dataTableDelez[,delezCol][(dataTableDelez[,delezCol] %in% delez)] <- 1	

+	}

+	

+	delezTable <- xtabs(dataTableDelez[,delezCol]~dataTableDelez[,1]+dataTableDelez[,2]+dataTableDelez[,3], dataTableDelez) / crosstabTable	

+}

+

+

+# vsako variablo (vrstico) pretvorimo v csv string

+vars <- sapply(vars, paste, collapse=",")

+

+# naredimo 2-d tabelo podatkov

+crosstabTable <- ftable(crosstabTable)

+crosstabTable <- sapply(crosstabTable, paste, collapse=", ")

+

+

+# podatke vrnemo v obliki var11,var12,var13..._var2...DATAfreq1_freq2...

+cat(paste(vars, sep="", collapse="_"))

+cat("--")

+cat(paste(crosstabTable, sep="", collapse="_"))

+cat("--")

+cat(paste(sumsVrstica, sep="", collapse="_"))

+cat("--")

+cat(paste(sumsStolpec, sep="", collapse="_"))

+cat("--")

+cat(paste(sums, sep="", collapse="_"))

+cat("--")

+cat(paste(x2, sep="", collapse="_"))

+

+# rezultati povprecja

+if(params[2] == '1'){

+	cat("--")

+	cat(paste(avgTable, sep="", collapse="_"))

+}

+

+# rezultati deleza

+if(length(params) > 2){

+	cat("--")

+	cat(paste(delezTable, sep="", collapse="_"))

+}
\ No newline at end of file
diff --git a/admin/survey/R/script/createCrosstabulation4.R b/admin/survey/R/script/createCrosstabulation4.R
new file mode 100644
index 0000000..6dd4784
--- /dev/null
+++ b/admin/survey/R/script/createCrosstabulation4.R
@@ -0,0 +1,108 @@
+# CROSSTABULACIJA 3 SPREMENLJIVK

+

+# preberemo parametre 1->folderpath

+params <- commandArgs(trailingOnly=TRUE)

+

+path <- params[1]

+

+# preberemo sfiltrirano tabelo s podatki

+dataTable <- read.table(

+	file = paste(path, 'admin/survey/R/TempData/crosstab_data.tmp', sep=''),

+	sep = ',', 

+	colClasses = 'numeric',

+	comment.char = '',

+	quote = '',

+	header = FALSE

+)

+

+	

+# iz tabele pobrisemo vrstice z vrednostmi manjsimi od 0 in sfiltriramo vse stolpce ki jih ne rabimo

+dataTableClean <- dataTable[apply(dataTable[,1:4], MARGIN = 1, function(x) all(x > -1)), 1:4]

+

+

+

+# izvedemo krostabulacijo

+crosstabTable <- xtabs(~dataTableClean[,1]+dataTableClean[,2]+dataTableClean[,3]+dataTableClean[,4], data=dataTableClean)

+

+

+

+# Vrednosti katerim pripadajo frekvence

+vars <- dimnames(crosstabTable)

+

+# izracunamo vsoto

+sumsVrstica <- apply(crosstabTable, c(3), sum)

+sumsStolpec <- apply(crosstabTable, c(4), sum)

+sums <- sum(crosstabTable)

+

+

+delezCol <- 5

+

+# POVPRECJE

+if(params[2] == '1'){

+	

+	#dataTableAvg <- dataTableClean[sapply(dataTableClean[,5], function(x) all(x > -1)), ]

+	dataTableAvg <- dataTableClean

+	dataTableAvg[,5][dataTableAvg[,5] < 0] <- 0

+	

+	avgTable <- xtabs(dataTableAvg[,5]~dataTableAvg[,1]+dataTableAvg[,2]+dataTableAvg[,3]+dataTableAvg[,4], dataTableAvg) / crosstabTable

+

+	delezCol <- delezCol + 1

+}

+

+

+# DELEZ

+if(length(params) > 2){

+

+	# array, ki vsebuje vrednosti stolpca, za katere racunamo delez

+	delez <- params[3]

+	delez <- unlist(strsplit(delez, ","))

+	delez <- sapply(delez, strtoi)

+	

+	# ce je -1 gre za checxbox stolpce ki imajo samo vrednost 1 in jih je vec

+	if(delez[1] == -1){

+		dataTableDelez <- dataTableClean

+		dataTableDelez[,delezCol][!(rowSums(dataTableDelez[delezCol:ncol(dataTableDelez)]) == ncol(dataTableDelez)-delezCol+1)] <- 0

+		dataTableDelez[,delezCol][(rowSums(dataTableDelez[delezCol:ncol(dataTableDelez)]) == ncol(dataTableDelez)-delezCol+1)] <- 1

+	}

+	else{

+		#dataTableDelez <- dataTableClean[sapply(dataTableClean[,delezCol], function(x) all(x %in% delez)), ]

+		dataTableDelez[,delezCol][!(dataTableDelez[,delezCol] %in% delez)] <- 0

+		dataTableDelez[,delezCol][(dataTableDelez[,delezCol] %in% delez)] <- 1

+	}

+	

+	delezTable <- xtabs(dataTableDelez[,delezCol]~dataTableDelez[,1]+dataTableDelez[,2]+dataTableDelez[,3]+dataTableDelez[,4], dataTableDelez) / crosstabTable	

+}

+

+

+# vsako variablo (vrstico) pretvorimo v csv string

+vars <- sapply(vars, paste, collapse=",")

+

+# naredimo 2-d tabelo podatkov

+crosstabTable <- ftable(crosstabTable)

+crosstabTable <- sapply(crosstabTable, paste, collapse=", ")

+

+

+# podatke vrnemo v obliki var11,var12,var13..._var2...DATAfreq1_freq2...

+cat(paste(vars, sep="", collapse="_"))

+cat("--")

+cat(paste(crosstabTable, sep="", collapse="_"))

+cat("--")

+cat(paste(sumsVrstica, sep="", collapse="_"))

+cat("--")

+cat(paste(sumsStolpec, sep="", collapse="_"))

+cat("--")

+cat(paste(sums, sep="", collapse="_"))

+cat("--")

+cat(paste(x2, sep="", collapse="_"))

+

+# rezultati povprecja

+if(params[2] == '1'){

+	cat("--")

+	cat(paste(avgTable, sep="", collapse="_"))

+}

+

+# rezultati deleza

+if(length(params) > 2){

+	cat("--")

+	cat(paste(delezTable, sep="", collapse="_"))

+}
\ No newline at end of file
diff --git a/admin/survey/R/script/createTTest.R b/admin/survey/R/script/createTTest.R
new file mode 100644
index 0000000..e3cd83f
--- /dev/null
+++ b/admin/survey/R/script/createTTest.R
@@ -0,0 +1,115 @@
+# preberemo parametre 1->folderpath

+params <- commandArgs(trailingOnly=TRUE)

+

+path <- params[1]

+

+# preberemo sfiltrirano tabelo s podatki

+dataTable <- read.table(

+	file = paste(path, 'admin/survey/R/TempData/ttest_data.tmp', sep=''),

+	sep = ',', 

+	colClasses = 'numeric',

+	comment.char = '',

+	quote = '',

+	header = FALSE

+)

+

+

+

+# ce ni checkbox moramo podatke preurediti (imamo 2 dodatna parametra)

+if(params[3] > 1){

+	

+	val1 <- params[2]

+	val2 <- params[3]

+

+	dataTableClean <- dataTable[sapply(dataTable[,1], function(x) all((x == val1) || (x == val2))), ]

+	

+	dataTable[,1][dataTable[,1] != val1] <- 0

+	dataTable[,1][dataTable[,1] == val1] <- 1

+	

+	dataTable[,2][dataTable[,2] != val2] <- 0

+	dataTable[,2][dataTable[,2] == val2] <- 1

+}

+

+

+# pocistimo vse vrstice, ki imajo za numeric (col 3) missing (< 0)

+dataTable <- dataTable[sapply(dataTable[,3], function(x) all(x > -1)), ]

+dataTableClean <- dataTableClean[sapply(dataTableClean[,3], function(x) all(x > -1)), ]

+

+

+# podmnozica vrednosti glede na 1. vrednost

+col1 <- subset(dataTable, dataTable[,1] == 1)

+# podmnozica vrednosti glede na 2. vrednost

+col2 <- subset(dataTable, dataTable[,2] == 1)

+

+

+# n

+n1 <- nrow(col1)

+n2 <- nrow(col2)

+

+# avg(x)

+avg1 <- mean(col1[,3])

+if(is.nan(avg1))

+	avg1 <- 0

+avg2 <- mean(col2[,3])

+if(is.nan(avg2))

+	avg2 <- 0

+

+# standardna deviacija (s^2??)

+sd1 <- sd(col1[,3])

+if(is.na(sd1))

+	sd1 <- 0

+sd2 <- sd(col2[,3])

+if(is.na(sd2))

+	sd2 <- 0

+

+# standardna napaka

+se1 = sd1 / sqrt(n1)

+if(is.nan(se1))

+	se1 <- 0

+se2 = sd2 / sqrt(n2)

+if(is.nan(se2))

+	se2 <- 0

+

+# kvadrat standardne napake

+se21 <- se1^2

+se22 <- se2^2

+

+# margini (1,96 * se)

+mar1 <- 1.96 * se1

+mar2 <- 1.96 * se2

+

+

+

+# izvedemo ttest (ce obstajata po vsaj dva primera vsake binarni vrednosti)

+if(n1 > 1 && n2 > 1){

+	ttest <- t.test(dataTableClean[,3]~dataTableClean[,1], var.equal=FALSE)

+

+	# razlika povprečij => $d = x1 -x2

+	d <- avg1 - avg2

+

+	# sed (std. error difference)

+	#sed <- sqrt(se21 + se22)

+	#sed <- ttest$estimate/ttest$statistic

+	sed <- (-1 * diff(ttest$estimate) / ttest$statistic)

+	

+	#T <- d / sed

+	T <- ttest$statistic

+	

+	# signifikanca

+	sig <- ttest$p.value

+	

+} else{

+	d <- 0

+	sed <- 0

+	T <- 0

+	sig <- 0

+}

+

+

+

+# podatke vrnemo v x1_x2_x3...--y1_y2...

+cat(paste(n1, avg1, sd1, se1, se21, mar1, sep="_"))

+cat("--")

+cat(paste(n2, avg2, sd2, se2, se22, mar2, sep="_"))

+cat("--")

+cat(paste(d, sed, T, sig, sep="_"))
\ No newline at end of file