Send output to:
Browser Blue - Charts White
Browser Black/White
CSV
Data:
255 280.2 299.9 339.2 374.2 393.5 389.2 381.7 375.2 369 357.4 352.1 346.5 342.9 340.3 328.3 322.9 314.3 308.9 294 285.6 281.2 280.3 278.8 274.5 270.4 263.4 259.9 258 262.7 284.7 311.3 322.1 327 331.3 333.3 321.4 327 320 314.7 316.7 314.4 321.3 318.2 307.2 301.3 287.5 277.7 274.4 258.8 253.3 251 248.4 249.5 246.1 244.5 243.6 244 240.8 249.8 248 259.4 260.5 260.8 261.3 259.5 256.6 257.9 256.5 254.2 253.3 253.8 255.5 257.1 257.3 253.2 252.8 252 250.7 252.2 250 251 253.4 251.2 255.6 261.1 258.9 259.9 261.2 264.7 267.1 266.4 267.7 268.6 267.5 268.5 268.5 270.5 270.9 270.1 269.3 269.8 270.1 264.9 263.7 264.8 263.7 255.9 276.2 360.1 380.5 373.7 369.8 366.6 359.3 345.8 326.2 324.5 328.1 327.5 324.4 316.5 310.9 301.5 291.7 290.4 287.4 277.7 281.6 288 276 272.9 283 283.3 276.8 284.5 282.7 281.2 287.4 283.1 284 285.5 289.2 292.5 296.4 305.2 303.9 311.5 316.3 316.7 322.5 317.1 309.8 303.8 290.3 293.7 291.7 296.5 289.1 288.5 293.8 297.7 305.4 302.7 302.5 303 294.5 294.1 294.5 297.1 289.4 292.4 287.9 286.6 280.5 272.4 269.2 270.6 267.3 262.5 266.8 268.8 263.1 261.2 266 262.5 265.2 261.3 253.7 249.2 239.1 236.4 235.2 245.2 246.2 247.7 251.4 253.3 254.8 250 249.3 241.5 243.3 248 253 252.9 251.5 251.6 253.5 259.8 334.1 448 445.8 445 448.2 438.2 439.8 423.4 410.8 408.4 406.7 405.9 402.7 405.1 399.6 386.5 381.4 375.2 357.7 359 355 352.7 344.4 343.8 338 339 333.3 334.4 328.3 330.7 330 331.6 351.2 389.4 410.9 442.8 462.8 466.9 461.7 439.2 430.3 416.1 402.5 397.3 403.3 395.9 387.8 378.6 377.1 370.4 362 350.3 348.2 344.6 343.5 342.8 347.6 346.6 349.5 342.1 342 342.8 339.3 348.2 333.7 334.7 354 367.7 363.3 358.4 353.1 343.1 344.6 344.4 333.9 331.7 324.3 321.2 322.4 321.7 320.5 312.8 309.7 315.6 309.7 304.6 302.5 301.5 298.8 291.3 293.6 294.6 285.9 297.6 301.1 293.8 297.7 292.9 292.1 287.2 288.2 283.8 299.9 292.4 293.3 300.8 293.7 293.1 294.4 292.1 291.9 282.5 277.9 287.5 289.2 285.6 293.2 290.8 283.1 275 287.8 287.8 287.4 284 277.8 277.6 304.9 294 300.9 324 332.9 341.6 333.4 348.2 344.7 344.7 329.3 323.5 323.2 317.4 330.1 329.2 334.9 315.8 315.4 319.6 317.3 313.8 315.8 311.3
Sample Range:
(leave blank to include all observations)
From:
To:
Chart options
R Code
x <-sort(x[!is.na(x)]) q1 <- function(data,n,p,i,f) { np <- n*p; i <<- floor(np) f <<- np - i qvalue <- (1-f)*data[i] + f*data[i+1] } q2 <- function(data,n,p,i,f) { np <- (n+1)*p i <<- floor(np) f <<- np - i qvalue <- (1-f)*data[i] + f*data[i+1] } q3 <- function(data,n,p,i,f) { np <- n*p i <<- floor(np) f <<- np - i if (f==0) { qvalue <- data[i] } else { qvalue <- data[i+1] } } q4 <- function(data,n,p,i,f) { np <- n*p i <<- floor(np) f <<- np - i if (f==0) { qvalue <- (data[i]+data[i+1])/2 } else { qvalue <- data[i+1] } } q5 <- function(data,n,p,i,f) { np <- (n-1)*p i <<- floor(np) f <<- np - i if (f==0) { qvalue <- data[i+1] } else { qvalue <- data[i+1] + f*(data[i+2]-data[i+1]) } } q6 <- function(data,n,p,i,f) { np <- n*p+0.5 i <<- floor(np) f <<- np - i qvalue <- data[i] } q7 <- function(data,n,p,i,f) { np <- (n+1)*p i <<- floor(np) f <<- np - i if (f==0) { qvalue <- data[i] } else { qvalue <- f*data[i] + (1-f)*data[i+1] } } q8 <- function(data,n,p,i,f) { np <- (n+1)*p i <<- floor(np) f <<- np - i if (f==0) { qvalue <- data[i] } else { if (f == 0.5) { qvalue <- (data[i]+data[i+1])/2 } else { if (f < 0.5) { qvalue <- data[i] } else { qvalue <- data[i+1] } } } } lx <- length(x) qval <- array(NA,dim=c(99,8)) mystep <- 25 mystart <- 25 if (lx>10){ mystep=10 mystart=10 } if (lx>20){ mystep=5 mystart=5 } if (lx>50){ mystep=2 mystart=2 } if (lx>=100){ mystep=1 mystart=1 } for (perc in seq(mystart,99,mystep)) { qval[perc,1] <- q1(x,lx,perc/100,i,f) qval[perc,2] <- q2(x,lx,perc/100,i,f) qval[perc,3] <- q3(x,lx,perc/100,i,f) qval[perc,4] <- q4(x,lx,perc/100,i,f) qval[perc,5] <- q5(x,lx,perc/100,i,f) qval[perc,6] <- q6(x,lx,perc/100,i,f) qval[perc,7] <- q7(x,lx,perc/100,i,f) qval[perc,8] <- q8(x,lx,perc/100,i,f) } bitmap(file='test1.png') myqqnorm <- qqnorm(x,col=2) qqline(x) grid() dev.off() load(file='createtable') a<-table.start() a<-table.row.start(a) a<-table.element(a,'Percentiles - Ungrouped Data',9,TRUE) a<-table.row.end(a) a<-table.row.start(a) a<-table.element(a,'p',1,TRUE) a<-table.element(a,hyperlink('http://www.xycoon.com/method_1.htm', 'Weighted Average at Xnp',''),1,TRUE) a<-table.element(a,hyperlink('http://www.xycoon.com/method_2.htm','Weighted Average at X(n+1)p',''),1,TRUE) a<-table.element(a,hyperlink('http://www.xycoon.com/method_3.htm','Empirical Distribution Function',''),1,TRUE) a<-table.element(a,hyperlink('http://www.xycoon.com/method_4.htm','Empirical Distribution Function - Averaging',''),1,TRUE) a<-table.element(a,hyperlink('http://www.xycoon.com/method_5.htm','Empirical Distribution Function - Interpolation',''),1,TRUE) a<-table.element(a,hyperlink('http://www.xycoon.com/method_6.htm','Closest Observation',''),1,TRUE) a<-table.element(a,hyperlink('http://www.xycoon.com/method_7.htm','True Basic - Statistics Graphics Toolkit',''),1,TRUE) a<-table.element(a,hyperlink('http://www.xycoon.com/method_8.htm','MS Excel (old versions)',''),1,TRUE) a<-table.row.end(a) for (perc in seq(mystart,99,mystep)) { a<-table.row.start(a) a<-table.element(a,round(perc/100,2),1,TRUE) for (j in 1:8) { a<-table.element(a,round(qval[perc,j],6)) } a<-table.row.end(a) } a<-table.end(a) table.save(a,file='mytable.tab')
Compute
Summary of computational transaction
Raw Input
view raw input (R code)
Raw Output
view raw output of R engine
Computing time
1 seconds
R Server
Big Analytics Cloud Computing Center
Click here to blog (archive) this computation