This is a test version of Biostars. For the public version, visit https://www.biostars.org.
Simple Question About Pvclust In R

Hey, im a beginner to R and trying to run pvclust so as to test a cluster solution.

I've managed to load data and run the heirachical cluster, however the code i find online for running pvclust is constantly producing errors - just wondering if someone can point out where I'm going wrong...

here is my code (data already transposed)

##loaddata

transpose <- 
  read.table("C:/Users/Tim/University/Advanced Design and Data Analysis/Assignment/transposed.csv",
   header=TRUE, sep=",", na.strings="NA", dec=".", strip.white=TRUE)

View(transpose)
hello <- hclust(dist(model.matrix(~-1 + 
  var001+var002+var003+var004+var005+var006+var007+var008+var009+var010+var011+var012+var013+var014+var015+var016+var017+var018+var019+var020+var021+var022+var023+var024+var025+var026+var027+var028+var029+var030+var031+var032+var033+var034+var035+var036+var037+var038+var039+var040+var041+var042+var043+var044+var045+var046+var047+var048+var049+var050+var051+var052+var053+var054+var055+var056+var057+var058+var059+var060+var061+var062+var063+var064+var065+var066+var067+var068+var069+var070+var071+var072+var073+var074+var075+var076+var077+var078+var079+var080+var081+var082+var083+var084+var085+var086+var087+var088+var089+var090+var091+var092+var093+var094+var095+var096+var097+var098+var099+var100+var101+var102+var103+var104+var105+var106+var107+var108+var109+var110+var111+var112+var113+var114+var115+var116+var117+var118+var119+var120+var121+var122+var123+var124+var125+var126+var127+var128+var129+var130+var131+var132+var133+var134+var135+var136+var137+var138+var139+var140+var141+var142+var143+var144+var145+var146+var147+var148+var149+var150+var151+var152+var153+var154+var155+var156+var157+var158+var159+var160+var161+var162+var163+var164+var165+var166+var167+var168+var169+var170+var171+var172+var173+var174+var175+var176,
   transpose)) , method= "ward")

plot(hello, main= "Cluster Dendrogram for Solution hello", xlab= 
  "Observation Number in Data Set transpose", sub="Method=ward; 
  Distance=euclidian")

###all the abiove works fine, except the bwlo where i try the pvclust

library(pvclust)

fit <- pvclust(transpose, method.hclust="ward",
   method.dist="euclidean")

plot(fit) 
pvrect(fit, alpha=.95)
###

the error comes back=


library(pvclust)
> fit <- pvclust(transpose, method.hclust="ward",
+    method.dist="euclidean")
Warning in dist(t(x), method) : NAs introduced by coercion
Error in hclust(distance, method = method.hclust) : 
  NA/NaN/Inf in foreign function call (arg 11)
> plot(fit) 
Error in plot(fit) : object 'fit' not found
> pvrect(fit, alpha=.95)
Error in nrow(x$edges) : object 'fit' not found
r
• 7,178 views • link
updated by Biostar Biostar • written by Max • Max

put your csv in a web accessible location

• 0 views • link
written by Jeremy Leipzig Jeremy Leipzig

1 answer

Basically, what the errors are telling you is this:

  1. When you run pvclust(), it calls a function named dist(). For some reason, this introduces NA (= missing values) into the data.
  2. pvclust() then uses the function hclust(), which fails because it cannot accept missing values.
  3. There is then no point running plot() or pvrect(), because the preceding operations failed to generate the object named 'fit', hence: object 'fit' not found.

So your problem is to figure out why "NAs introduced by coercion" is occurring. It's probably because the input data (transpose) is not in the correct form.

Can't provide much more help than that without seeing the CSV file. I suggest reading the package documentation (difficult for beginners, I know) to try and understand how each function works and what arguments it expects. For example, pvclust comes with some sample data which you can load using data(lung), examine to see how it is structured and run through some of the package functions. See the pvclust PDF.

Copy/paste from the Web is rarely a good idea when you don't understand the underlying processes.

• 0 views • link
written by Neilfws Neilfws

thanks you were right, the data needed to be represented as a table i.e., t(x). thanks, fixed

• 0 views • link
written by Max • Max

You're welcome. Votes for answers are appreciated.

• 0 views • link
written by Neilfws Neilfws

Log in to answer this question.

More posts like this