/archive/source.archive/bm.and.ou/rjmcmc.12042010.R

http://github.com/eastman/auteur · R · 1310 lines · 1117 code · 140 blank · 53 comment · 222 complexity · 535a2e3626d1e9d9ed5c52c6c27f4d54 MD5 · raw file 

    

  1. ## Written by LJ HARMON (2008), AL HIPP (2008), and JM EASTMAN (2010)

    2. 

  2. # MODELING TRAIT EVOLUTION by BROWNIAN MOTION

    3. 

  3. 

    4. 

  4. rjMCMC.bm<-function (phy, data, ngen=1000, sampleFreq=100, probMergeSplit=0.05, probRoot=0.01, lambdaK=log(2), constrainK=FALSE, jumpsize=NULL, fileBase="result", simplestart=FALSE, switcheroo=TRUE) 

    5. 

  5. { 

    6. 

  6. # ngen=1000; sampleFreq=100; probMergeSplit=0.05; probRoot=0.01; lambdaK=log(2); constrainK=FALSE; jumpsize=NULL; fileBase="result"; simplestart=FALSE; switcheroo=TRUE 

    7. 

  7. 	model="BM"

    8. 

  8. 	heat=1

    9. 

  9. 	if(is.null(jumpsize)) jumpsize=log(2)

    10. 

  10. 	require(geiger)

    11. 

  11. 		

    12. 

  12. ### prepare data for rjMCMC

    13. 

  13. 	cur.model		<- model

    14. 

  14. 

    15. 

  15. 	dataList		<- prepare.data(phy, data)			

    16. 

  16. 	ape.tre			<- dataList$ape.tre			

    17. 

  17. 	orig.dat		<- dataList$orig.dat						

    18. 

  18. 	nn				<- length(ape.tre$edge.length)	

    19. 

  19. 	node.des		<- sapply(unique(c(ape.tre$edge[1,1],ape.tre$edge[,2])), function(x) get.descendants.of.node(x, ape.tre))

    20. 

  20. 	names(node.des) <- c(ape.tre$edge[1,1], unique(ape.tre$edge[,2]))

    21. 

  21. 	subtrees		<- subtrees(phy)

    22. 

  22. 	subtrees.marker <- sapply(subtrees, function(x) return(min(x$node.label)))

    23. 

  23. 

    24. 

  24. # initialize parameters

    25. 

  25. 	if(is.numeric(constrainK) & (constrainK > length(ape.tre$edge) | constrainK < 1)) stop("Constraint on rate shifts is nonsensical. Ensure that constrainK is at least 1 and less than the number of available nodes in the tree.")

    26. 

  26. 

    27. 

  27. 	if(simplestart | is.numeric(constrainK) ) {

    28. 

  28. 		if(is.numeric(constrainK)) {

    29. 

  29. 			init.rate	<- generate.starting.point(orig.dat, ape.tre, node.des, theta=FALSE, K=constrainK, jumpsize=jumpsize)

    30. 

  30. 		} else {

    31. 

  31. 			init.rate	<- list(values=rep(fitContinuous(ape.tre,orig.dat)$Trait1$beta,length(ape.tre$edge.length)),delta=rep(0,length(ape.tre$edge.length)))

    32. 

  32. 		}

    33. 

  33. 	} else {

    34. 

  34. 		init.rate	<- generate.starting.point(orig.dat, ape.tre, node.des, theta=FALSE, K=constrainK, jumpsize=jumpsize )

    35. 

  35. 	}

    36. 

  36. 		

    37. 

  37.     cur.rates		<- init.rate$values			

    38. 

  38. 	cur.delta.rates	<- init.rate$delta

    39. 

  39. 	cur.root		<- adjust.value(mean(orig.dat), jumpsize)

    40. 

  40. #	design.vcv		<- design.vcv.phylo(ape.tre)

    41. 

  41. 	cur.vcv			<- update.vcv(ape.tre, cur.rates)

    42. 

  42. 	

    43. 

  43. 	mod.cur = new.bm.lik.fn(cur.rates, cur.root, orig.dat, cur.vcv)

    44. 

  44. 	

    45. 

  45. # proposal counts

    46. 

  46. 	nRateProp =		0	

    47. 

  47. 	nRateSwapProp = 0									

    48. 

  48. 	nRcatProp =		0										

    49. 

  49. 	nRootProp =		0

    50. 

  50. 	nRateOK =		0											

    51. 

  51. 	nRateSwapOK =	0

    52. 

  52. 	nRcatOK =		0											

    53. 

  53. 	nRootOK =		0

    54. 

  54. 

    55. 

  55. 	cOK=c(												

    56. 

  56. 		nRateOK,

    57. 

  57. 		nRateSwapOK,

    58. 

  58. 		nRcatOK, 

    59. 

  59. 		nRootOK 

    60. 

  60. 	)

    61. 

  61. 		

    62. 

  62. 	tickerFreq=ceiling(ngen/30)

    63. 

  63. 	

    64. 

  64. # file handling

    65. 

  65. 	parmBase=paste(model, fileBase, "parameters/",sep=".")

    66. 

  66. 	if(!file.exists(parmBase)) dir.create(parmBase)

    67. 

  67. 	parlogger(model=model, init=TRUE, node.des, parmBase=parmBase)

    68. 

  68. 	errorLog=file(paste(parmBase,paste(cur.model, fileBase, "rjTraitErrors.log",sep="."),sep="/"),open='w+')

    69. 

  69. 	generate.error.message(i=NULL, mod.cur=NULL, mod.new=NULL, lnR=NULL, errorLog=errorLog, init=TRUE)

    70. 

  70. 	runLog=file(paste(parmBase,paste(cur.model, fileBase, "rjTrait.log",sep="."),sep="/"),open='w+')

    71. 

  71. 	generate.log(bundled.parms=NULL, cur.model, file=runLog, init=TRUE)

    72. 

  72. 

    73. 

  73. 

    74. 

  74. ### Begin rjMCMC

    75. 

  75.     for (i in 1:ngen) {

    76. 

  76.         lnLikelihoodRatio <- lnHastingsRatio <- lnPriorRatio <- 0

    77. 

  77. 		startparms = c(nRateProp, nRateSwapProp, nRcatProp, nRootProp)

    78. 

  78. 

    79. 

  79. ## BM IMPLEMENTATION ##

    80. 

  80. 		while(1) {

    81. 

  81. 			if (runif(1) < (2 * probMergeSplit) & !constrainK) {			# adjust rate categories

    82. 

  82. 				nr=split.or.merge(cur.delta.rates, cur.rates, ape.tre, node.des, lambdaK, theta=FALSE)

    83. 

  83. 				new.rates=nr$new.values

    84. 

  84. 				new.delta.rates=nr$new.delta

    85. 

  85. 				new.root=cur.root

    86. 

  86. 				nRcatProp=nRcatProp+1

    87. 

  87. 				lnHastingsRatio=nr$lnHastingsRatio

    88. 

  88. 				lnPriorRatio=nr$lnPriorRatio

    89. 

  89. 				break()

    90. 

  90. 			} else { 

    91. 

  91. 				if(runif(1)<probRoot) {										# adjust root

    92. 

  92. 					new.root=adjust.value(cur.root, jumpsize)

    93. 

  93. 					new.rates=cur.rates

    94. 

  94. 					new.delta.rates=cur.delta.rates

    95. 

  95. 					nRootProp=nRootProp+1

    96. 

  96. 					break()

    97. 

  97. 				} else {													

    98. 

  98. 					if(runif(1)>0.1 & length(unique(cur.rates))>1 & switcheroo) {		# swap rates classes

    99. 

  99. 						nr=switcheroo(cur.rates)

    100. 

  100. 						new.rates=nr$new.values ##

    101. 

  101. 						new.delta.rates=cur.delta.rates 

    102. 

  102. 						new.root=cur.root

    103. 

  103. 						nRateSwapProp=nRateSwapProp+1

    104. 

  104. 						break()

    105. 

  105. 					} else {												# adjust rates

    106. 

  106. 						new.rates=adjust.rate(cur.rates, jumpsize)

    107. 

  107. 						new.delta.rates=cur.delta.rates

    108. 

  108. 						new.root=cur.root

    109. 

  109. 						nRateProp=nRateProp+1

    110. 

  110. 						break()

    111. 

  111. 					} 

    112. 

  112. 				}

    113. 

  113. 			}

    114. 

  114. 		}

    115. 

  115. 		

    116. 

  116. 		if(any(new.rates!=cur.rates)) new.vcv=update.vcv(ape.tre, new.rates) else new.vcv=cur.vcv

    117. 

  117. 		mod.new=try(new.bm.lik.fn(new.rates, new.root, orig.dat, new.vcv), silent=TRUE)

    118. 

  118. 		

    119. 

  119. 		if(inherits(mod.new, "try-error")) {mod.new=as.list(mod.new); mod.new$lnL=Inf}

    120. 

  120. 		if(!is.infinite(mod.new$lnL)) {

    121. 

  121. 			lnLikelihoodRatio = mod.new$lnL - mod.cur$lnL

    122. 

  122. 		} else {

    123. 

  123. 			lnLikelihoodRatio = -Inf

    124. 

  124. 			failures=paste("./failed", model, "parameters/",sep=".")

    125. 

  125. 			if(!file.exists(failures)) dir.create(failures)

    126. 

  126. 			write.table(matrix(c(i, new.rates),nrow=1),file=paste(failures,"failed.rates.txt",sep="/"),append=TRUE,col=FALSE,row=FALSE,quote=FALSE)

    127. 

  127. 			write.table(matrix(c(i, new.root),nrow=1),file=paste(failures,"failed.root.txt",sep="/"),append=TRUE,col=FALSE,row=FALSE,quote=FALSE)

    128. 

  128. 		}

    129. 

  129. 		

    130. 

  130. 	# compare likelihoods

    131. 

  131. 		endparms = c(nRateProp=nRateProp, nRateSwapProp=nRateSwapProp, nRcatProp=nRcatProp, nRootProp=nRootProp)

    132. 

  132. 

    133. 

  133. 		r=assess.lnR((heat * lnLikelihoodRatio + heat * lnPriorRatio + lnHastingsRatio)->lnR)

    134. 

  134. 

    135. 

  135. 		if (runif(1) <= r$r) {			# adopt proposal

    136. 

  136. 			decision="adopt"

    137. 

  137. 			cur.root <- new.root

    138. 

  138. 			cur.rates <- new.rates

    139. 

  139. 			cur.delta.rates <- new.delta.rates

    140. 

  140. 			mod.cur <- mod.new

    141. 

  141. 			cur.vcv <- new.vcv

    142. 

  142. 			curr.lnL <- mod.new$lnL

    143. 

  143. 			cOK <- determine.accepted.proposal(startparms, endparms, cOK)

    144. 

  144. 		} else {						# deny proposal	

    145. 

  145. 			decision="reject"

    146. 

  146. 			curr.lnL <- mod.cur$lnL 

    147. 

  147. 		}

    148. 

  148. 		

    149. 

  149. 		if(is.infinite(curr.lnL)) stop("starting point has exceptionally poor likelihood")

    150. 

  150. 

    151. 

  151. 		if(r$error) generate.error.message(i, mod.cur, mod.new, lnR, errorLog)

    152. 

  152. 		

    153. 

  153. 		if(i%%tickerFreq==0) {

    154. 

  154. 			if(i==tickerFreq) cat("|",rep(" ",9),toupper("generations complete"),rep(" ",9),"|","\n")

    155. 

  155. 			cat(". ")

    156. 

  156. 		}

    157. 

  157. 		

    158. 

  158. 		if(i%%sampleFreq==0) {

    159. 

  159. 			bundled.parms=list(gen=i, mrate=exp(mean(log(cur.rates))), cats=sum(cur.delta.rates)+1, root=cur.root, lnL=curr.lnL)

    160. 

  160. 			generate.log(bundled.parms, cur.model, file=runLog)			

    161. 

  161. 			parlogger(model=model, init=FALSE, node.des=node.des, i=i, curr.lnL=curr.lnL, cur.root=cur.root, cur.rates=cur.rates, cur.delta.rates=cur.delta.rates, parmBase=parmBase)

    162. 

  162. 		}

    163. 

  163. 	}

    164. 

  164. 	

    165. 

  165. # End rjMCMC

    166. 

  166. 	close(errorLog)

    167. 

  167. 	close(runLog)

    168. 

  168. 	summarize.run(cOK, endparms, cur.model)	

    169. 

  169. }

    170. 

  170. 

    171. 

  171. rjMCMC.ou<-function (phy, data, ngen=1000, sampleFreq=100, probMergeSplit=0.05, probRoot=0.01, probRate=0.1, probAlpha=0.01, upperAlpha=100, lambdaK=log(2), constrainK=FALSE, jumpsize=2, fileBase="result", simplestart=FALSE, switcheroo=TRUE) 

    172. 

  172. {

    173. 

  173. #	load("rjMCMC/versions/fakedata.rda"); phy=sims$phy; data=sims[[2]]; ngen=1000; sampleFreq=100; probMergeSplit=0.05; probRoot=0.01; probRate=0.1; probAlpha=0.01; lambdaK=log(2); constrainK=FALSE; jumpsize=2; upperAlpha=100; fileBase="result"; simplestart=FALSE 

    174. 

  174. 	

    175. 

  175. 	model="OU"

    176. 

  176. 	heat=1

    177. 

  177. 	require(ouch)

    178. 

  178. 	require(geiger)

    179. 

  179. 	abs.low.alpha=1e-13

    180. 

  180. 	if(length(upperAlpha)==2 & upperAlpha[1]>abs.low.alpha) {

    181. 

  181. 		alpha.bounds=upperAlpha

    182. 

  182. 	} else if(!is.null(upperAlpha) & is.numeric(upperAlpha)){

    183. 

  183. 		alpha.bounds=c(abs.low.alpha, upperAlpha)

    184. 

  184. 	} else {

    185. 

  185. 		alpha.bounds=c(abs.low.alpha, 100)

    186. 

  186. 	}

    187. 

  187. 	

    188. 

  188. ### prepare data for rjMCMC

    189. 

  189. 	cur.model		<- model								

    190. 

  190. 	

    191. 

  191. 	dataList		<- prepare.ouchdata(phy, data)			

    192. 

  192. 	ape.tre			<- dataList$ape.tre			

    193. 

  193. 	orig.dat		<- dataList$orig.dat	

    194. 

  194. 	ouch.tre		<- dataList$ouch.tre

    195. 

  195. 	ouch.dat		<- dataList$ouch.dat

    196. 

  196. 	ape.to.ouch.order <- match(as.numeric(names(dataList$ape.ordered[-1])), as.numeric(ape.tre$edge[,2]))

    197. 

  197. 	nn				<- length(ape.tre$edge.length)	

    198. 

  198. 	node.des		<- sapply(unique(c(ape.tre$edge[1,1],ape.tre$edge[,2])), function(x) get.descendants.of.node(x, ape.tre))

    199. 

  199. 	names(node.des) <- c(ape.tre$edge[1,1], unique(ape.tre$edge[,2]))

    200. 

  200. 	subtrees		<- subtrees(phy)

    201. 

  201. 	subtrees.marker <- sapply(subtrees, function(x) return(min(x$node.label)))

    202. 

  202. 	

    203. 

  203. # initialize parameters

    204. 

  204. 	if(is.numeric(constrainK) & (constrainK > length(ape.tre$edge) | constrainK < 1)) stop("Constraint on rate shifts is nonsensical. Ensure that constrainK is at least 1 and less than the number of available nodes in the tree.")

    205. 

  205. 	

    206. 

  206. 	if(simplestart | is.numeric(constrainK) ) {

    207. 

  207. 		if(is.numeric(constrainK)) {

    208. 

  208. 			init.theta	<- generate.starting.point(orig.dat, ape.tre, node.des, theta=TRUE, K=constrainK, jumpsize=jumpsize)

    209. 

  209. 		} else {

    210. 

  210. 			init.theta	<- list(values=rep(mean(orig.dat),length(ape.tre$edge.length)),delta=rep(0,length(ape.tre$edge.length)))

    211. 

  211. 		}

    212. 

  212. 	} else {

    213. 

  213. 		init.theta	<- generate.starting.point(orig.dat, ape.tre, node.des, theta=TRUE, K=constrainK, jumpsize=jumpsize )

    214. 

  214. 	}

    215. 

  215. 	

    216. 

  216. 	cur.rate		<- fitContinuous(ape.tre,orig.dat)$Trait1$beta

    217. 

  217. 	cur.alpha		<- min(runif(1000, alpha.bounds[1], alpha.bounds[2]))

    218. 

  218.     cur.theta		<- init.theta$values			

    219. 

  219. 	cur.delta.theta	<- init.theta$delta

    220. 

  220. 	cur.root.theta	<- assign.root.theta(cur.theta,ape.tre)

    221. 

  221. 	cur.regimes		<- c(cur.root.theta,cur.theta) 

    222. 

  222. 	

    223. 

  223. 	mod.cur = new.ou.lik.fn(cur.rate, cur.alpha, cur.regimes, cur.theta, ouch.dat, ouch.tre, ape.to.ouch.order)

    224. 

  224. 	

    225. 

  225. # proposal counts

    226. 

  226. 	nRateProp =		0	

    227. 

  227. 	nRootProp =		0

    228. 

  228. 	nAlphaProp =	0										

    229. 

  229. 	nTcatProp =		0										

    230. 

  230. 	nThetaProp =	0

    231. 

  231. 	nRateOK =		0											

    232. 

  232. 	nRootOK =		0

    233. 

  233. 	nAlphaOK =		0										

    234. 

  234. 	nTcatOK =		0

    235. 

  235. 	nThetaOK =		0

    236. 

  236. 	

    237. 

  237. 	cOK=c(												

    238. 

  238. 		  nRateOK,

    239. 

  239. 		  nRootOK,

    240. 

  240. 		  nAlphaOK,

    241. 

  241. 		  nTcatOK,

    242. 

  242. 		  nThetaOK

    243. 

  243. 		  )

    244. 

  244. 	

    245. 

  245. 	tickerFreq=ceiling(ngen/30)

    246. 

  246. 	

    247. 

  247. # file handling

    248. 

  248. 	parmBase=paste(model, fileBase, "parameters/",sep=".")

    249. 

  249. 	if(!file.exists(parmBase)) dir.create(parmBase)

    250. 

  250. 	parlogger(model=model, init=TRUE, node.des, parmBase=parmBase)

    251. 

  251. 	errorLog=file(paste(parmBase,paste(cur.model, fileBase, "rjTraitErrors.log",sep="."),sep="/"),open='w+')

    252. 

  252. 	generate.error.message(i=NULL, mod.cur=NULL, mod.new=NULL, lnR=NULL, errorLog=errorLog, init=TRUE)

    253. 

  253. 	runLog=file(paste(parmBase,paste(cur.model, fileBase, "rjTrait.log",sep="."),sep="/"),open='w+')

    254. 

  254. 	generate.log(bundled.parms=NULL, cur.model, file=runLog, init=TRUE)

    255. 

  255. 	

    256. 

  256. 	probSwap=max(c(probRoot, probRate, probAlpha))

    257. 

  257. 	minorprobs=sum(c(probRoot, probRate, probAlpha, probSwap))

    258. 

  258. ### Begin rjMCMC

    259. 

  259.     for (i in 1:ngen) {

    260. 

  260. 		

    261. 

  261. 		# loop to ensure that each state produces a sensible proposal 

    262. 

  262. 		while(1) {

    263. 

  263. 			lnLikelihoodRatio <- lnHastingsRatio <- lnPriorRatio <- 0

    264. 

  264. 			startparms = c(nRateProp, nRootProp, nAlphaProp, nTcatProp, nThetaProp)

    265. 

  265. 			

    266. 

  266. 			## OU IMPLEMENTATION ##

    267. 

  267. 			if (runif(1) < (2 * probMergeSplit) & !constrainK) {									# adjust theta categories

    268. 

  268. 				nr=split.or.merge(cur.delta.theta, cur.theta, ape.tre, node.des, lambdaK, theta=TRUE)

    269. 

  269. 				new.alpha=cur.alpha

    270. 

  270. 				new.theta=nr$new.values ##

    271. 

  271. 				new.delta.theta=nr$new.delta ##

    272. 

  272. 				new.root.theta<-nrt<-assign.root.theta(new.theta, ape.tre) ## 

    273. 

  273. 				new.regimes=c(nrt,new.theta) ##

    274. 

  274. 				new.rate=cur.rate 

    275. 

  275. 				nTcatProp=nTcatProp+1

    276. 

  276. 				lnHastingsRatio=nr$lnHastingsRatio

    277. 

  277. 				lnPriorRatio=nr$lnPriorRatio

    278. 

  278. 				

    279. 

  279. 			} else { # neither split nor merge

    280. 

  280. 				while(1) {

    281. 

  281. 					if(runif(1)<(probAlpha/minorprobs)) {											# adjust alpha

    282. 

  282. 						new.alpha=adjust.rate(cur.alpha, jumpsize) ##

    283. 

  283. 						new.root.theta=cur.root.theta

    284. 

  284. 						new.theta=cur.theta

    285. 

  285. 						new.delta.theta=cur.delta.theta

    286. 

  286. 						new.regimes=cur.regimes 

    287. 

  287. 						new.rate=cur.rate

    288. 

  288. 						nAlphaProp=nAlphaProp+1

    289. 

  289. 						break()

    290. 

  290. 					} else if(runif(1)<(probRoot/minorprobs)) {										# adjust root

    291. 

  291. 						new.alpha=cur.alpha									

    292. 

  292. 						new.root.theta=assign.root.theta(cur.theta, ape.tre) ##

    293. 

  293. 						new.theta=cur.theta

    294. 

  294. 						new.delta.theta=cur.delta.theta

    295. 

  295. 						new.regimes=cur.regimes 

    296. 

  296. 						new.rate=cur.rate 

    297. 

  297. 						nRootProp=nRootProp+1					

    298. 

  298. 						break()

    299. 

  299. 					} else if(runif(1)<(probRate/minorprobs)) {										# adjust rate

    300. 

  300. 						new.alpha=cur.alpha									

    301. 

  301. 						new.root.theta=cur.root.theta

    302. 

  302. 						new.theta=cur.theta

    303. 

  303. 						new.delta.theta=cur.delta.theta

    304. 

  304. 						new.regimes=cur.regimes 

    305. 

  305. 						new.rate=adjust.rate(cur.rate, jumpsize) ##

    306. 

  306. 						nRateProp=nRateProp+1

    307. 

  307. 						break()

    308. 

  308. 					} else if(runif(1)<0.5*(probSwap/minorprobs) & length(unique(cur.theta))>1 & switcheroo) {	# swap thetas

    309. 

  309. 						nr=switcheroo(cur.theta)

    310. 

  310. 						new.alpha=cur.alpha 

    311. 

  311. 						new.theta=nr$new.values ##

    312. 

  312. 						new.delta.theta=cur.delta.theta 

    313. 

  313. 						new.root.theta<-nrt<-assign.root.theta(new.theta, ape.tre) ## 

    314. 

  314. 						new.regimes=c(nrt,new.theta) ##

    315. 

  315. 						new.rate=cur.rate

    316. 

  316. 						nThetaProp=nThetaProp+1					

    317. 

  317. 						break()

    318. 

  318. 					} else {																		# adjust thetas

    319. 

  319. 						new.alpha=cur.alpha 

    320. 

  320. 						new.theta=adjust.value(cur.theta, jumpsize, theta=TRUE) ##

    321. 

  321. 						new.delta.theta=cur.delta.theta 

    322. 

  322. 						new.root.theta<-nrt<-assign.root.theta(new.theta, ape.tre) ## 

    323. 

  323. 						new.regimes=c(nrt,new.theta) ##

    324. 

  324. 						new.rate=cur.rate

    325. 

  325. 						nThetaProp=nThetaProp+1					

    326. 

  326. 						break()

    327. 

  327. 					}

    328. 

  328. 					

    329. 

  329. 				}

    330. 

  330. 			}

    331. 

  331. 			if(within.range(new.alpha, alpha.bounds[1], alpha.bounds[2])) {

    332. 

  332. 				mod.new=try(new.ou.lik.fn(new.rate, new.alpha, new.regimes, new.theta, ouch.dat, ouch.tre, ape.to.ouch.order),silent=TRUE)

    333. 

  333. 				if(!inherits(mod.new, "try-error")) break()

    334. 

  334. 			}

    335. 

  335. 

    336. 

  336. 		}

    337. 

  337. 		lnLikelihoodRatio = mod.new$lnL - mod.cur$lnL

    338. 

  338. 	

    339. 

  339. 		

    340. 

  340. # compare likelihoods

    341. 

  341. 		endparms = c(nRateProp, nRootProp, nAlphaProp, nTcatProp, nThetaProp)

    342. 

  342. 

    343. 

  343. 		r=assess.lnR((heat * lnLikelihoodRatio + heat * lnPriorRatio + lnHastingsRatio)->lnR)

    344. 

  344. 		

    345. 

  345. 		if (runif(1) <= r$r) {			# adopt proposal

    346. 

  346. 			decision="adopt"

    347. 

  347. 			cur.alpha <- new.alpha

    348. 

  348. 			cur.theta <- new.theta

    349. 

  349. 			cur.delta.theta <- new.delta.theta

    350. 

  350. 			cur.root.theta <- new.root.theta

    351. 

  351. 			cur.regimes <- new.regimes

    352. 

  352. 			cur.rate <- new.rate

    353. 

  353. 			cOK <- determine.accepted.proposal(startparms, endparms, cOK)

    354. 

  354. 			mod.cur <- mod.new

    355. 

  355. 			curr.lnL <- mod.new$lnL

    356. 

  356. 		} else {						# deny proposal	

    357. 

  357. 			decision="reject"

    358. 

  358. 			curr.lnL <- mod.cur$lnL 

    359. 

  359. 		}

    360. 

  360. 		

    361. 

  361. 		if(is.infinite(curr.lnL)) stop("starting point has exceptionally poor likelihood")

    362. 

  362. 		

    363. 

  363. 		if(r$error) generate.error.message(i, mod.cur, mod.new, lnR, errorLog)

    364. 

  364. 		

    365. 

  365. 		if(i%%tickerFreq==0) {

    366. 

  366. 			if(i==tickerFreq) cat("|",rep(" ",9),toupper("generations complete"),rep(" ",9),"|","\n")

    367. 

  367. 			cat(". ")

    368. 

  368. 		}

    369. 

  369. 		

    370. 

  370. 		if(i%%sampleFreq==0) {

    371. 

  371. 			bundled.parms=list(gen=i, rate=cur.rate, regimes=sum(cur.delta.theta)+1, root=cur.root.theta, mean.theta=mean(cur.theta), alpha=cur.alpha, lnL=curr.lnL)

    372. 

  372. 			generate.log(bundled.parms, cur.model, file=runLog)			

    373. 

  373. 			parlogger(model=model, init=FALSE, node.des=node.des, i=i, curr.lnL=curr.lnL, cur.root=cur.root.theta, cur.rates=cur.rate, cur.alpha=cur.alpha, cur.theta=cur.theta, cur.delta.theta=cur.delta.theta, parmBase=parmBase)

    374. 

  374. 		}

    375. 

  375. 	}

    376. 

  376. 	

    377. 

  377. # End rjMCMC

    378. 

  378. 	close(errorLog)

    379. 

  379. 	close(runLog)

    380. 

  380. 	summarize.run(cOK, endparms, cur.model)	

    381. 

  381. }

    382. 

  382. 

    383. 

  383. 

    384. 

  384. ## AUXILIARY FUNCTIONS

    385. 

  385. 

    386. 

  386. new.ou.lik.fn <- function(rate, alpha, regimes, theta, ouch.dat, ouch.tre, ape.to.ouch.order) { # from OUCH 2.7-1:::hansen()

    387. 

  387. # mod 11.17.2010 JM Eastman; checked -- returns very similar if tested against fitted values from hansen()

    388. 

  388. #	rate=new.rate; alpha=new.alpha; regimes=new.regimes; theta=new.theta

    389. 

  389. 	ouch.dat$regimes=as.factor(regimes[c(1,ape.to.ouch.order+1)])

    390. 

  390. #	theta=sort(unique(theta))

    391. 

  391. 	alpha.ouch=as.matrix(alpha)

    392. 

  392. 	n <- length(which(!is.na(ouch.dat['trait'])))	

    393. 

  393. 	beta=ouch:::regime.spec(ouch.tre, ouch.dat['regimes']) 

    394. 

  394. 	theta=as.numeric(levels(ouch:::sets.of.regimes(ouch.tre,ouch.dat['regimes'])$regimes))

    395. 

  395. 	dat=ouch.dat$trait[n:nrow(ouch.dat)]

    396. 

  396. 	names(dat)=ouch.dat$labels[n:nrow(ouch.dat)]

    397. 

  397. 	ev <- eigen(alpha.ouch,symmetric=TRUE)

    398. 

  398. 	w <- .Call(ouch:::ouch_weights,object=ouch.tre,lambda=ev$values,S=ev$vectors,beta=beta) 

    399. 

  399. 	v <- .Call(ouch:::ouch_covar,object=ouch.tre,lambda=ev$values,S=ev$vectors,sigma.sq=rate)

    400. 

  400. 	y <- matrix(theta,ncol=1)

    401. 

  401. 	e <- w%*%y-dat

    402. 

  402. 	dim(y) <- dim(y)[1]

    403. 

  403. 	dim(e) <- n

    404. 

  404. 	q <- e%*%solve(v,e)

    405. 

  405. 	det.v <- determinant(v,logarithm=TRUE)

    406. 

  406. 	dev <- n*log(2*pi)+as.numeric(det.v$modulus)+q[1,1]

    407. 

  407. 	list(

    408. 

  408. 		 theta=theta,

    409. 

  409. 		 weight=w,

    410. 

  410. 		 vcov=v,

    411. 

  411. 		 resids=e,

    412. 

  412. 		 lnL=-0.5*dev

    413. 

  413. 		 )

    414. 

  414. }

    415. 

  415. 

    416. 

  416. new.bm.lik.fn <- function(rates, root, orig.dat, vcv) { # from LJ HARMON	

    417. 

  417. # mod 12.02.2010 JM Eastman: using determinant()$modulus rather than det() to stay in log-space

    418. 

  418. 	y=orig.dat

    419. 

  419. 	

    420. 

  420. 	b <- vcv

    421. 

  421. 	w <- rep(root, nrow(b))

    422. 

  422. 	num <- -t(y-w)%*%solve(b)%*%(y-w)/2

    423. 

  423. 	den <- 0.5*(length(y)*log(2*pi) + as.numeric(determinant(b)$modulus))

    424. 

  424. 	list(

    425. 

  425. 		 root=root,

    426. 

  426. 		 lnL = (num-den)[1,1]

    427. 

  427. 		 )	

    428. 

  428. }

    429. 

  429. 

    430. 

  430. split.or.merge <- function(cur.delta, cur.values, phy, node.des, lambda=lambda, theta=FALSE) { 

    431. 

  431. # mod 12.05.2010 JM Eastman

    432. 

  432. #	cur.delta=cur.delta.rates; cur.values=cur.rates; theta=FALSE; lambda=log(2); jumpsize=2

    433. 

  433. 	bb=cur.delta

    434. 

  434. 	vv=cur.values

    435. 

  435. 	names(vv)<-names(bb)<-phy$edge[,2]

    436. 

  436. 	new.bb=choose.one(bb)

    437. 

  437. 	new.vv=vv

    438. 

  438. 	

    439. 

  439. 	s=names(new.bb[bb!=new.bb])

    440. 

  440. 	all.shifts=as.numeric(names(bb[bb>0]))

    441. 

  441. 	all.D=node.des[[which(names(node.des)==s)]]

    442. 

  442. 	if(length(all.D)!=0) {

    443. 

  443. 		untouchable=unlist(lapply(all.shifts[all.shifts>s], function(x)node.des[[which(names(node.des)==x)]]))

    444. 

  444. 		remain.unchanged=union(all.shifts, untouchable)

    445. 

  445. 	} else {

    446. 

  446. 		untouchable=NULL

    447. 

  447. 		remain.unchanged=list()

    448. 

  448. 	}

    449. 

  449. 	

    450. 

  450. 	marker=match(s, names(new.vv))

    451. 

  451. 	nn=length(vv)

    452. 

  452. 	K=sum(bb)+1

    453. 

  453. 	N=Ntip(phy)

    454. 

  454. 	

    455. 

  455. 	ncat=sum(bb)

    456. 

  456. 	cur.vv=as.numeric(vv[marker])

    457. 

  457. 	ca.vv=length(which(vv==cur.vv))

    458. 

  458. 	

    459. 

  459. 	if(sum(new.bb)>sum(bb)) {			# add transition: SPLIT

    460. 

  460. #		print("s")

    461. 

  461. 		decision="split"

    462. 

  462. 		n.desc=sum(!all.D%in%remain.unchanged)+1

    463. 

  463. 		n.split=sum(vv==cur.vv)-n.desc

    464. 

  464. 		if(theta) u=split.value(cur.vv=cur.vv, n.desc=n.desc, n.split=n.split, factor=lambda) else u=split.rate(cur.vv=cur.vv, n.desc=n.desc, n.split=n.split, factor=lambda)

    465. 

  465. 		nr.split=u$nr.split

    466. 

  466. 		nr.desc=u$nr.desc

    467. 

  467. 		new.vv[vv==cur.vv]=nr.split

    468. 

  468. 		if(length(remain.unchanged)==0) {	# assign new value to all descendants

    469. 

  469. 			new.vv[match(all.D,names(new.vv))] = nr.desc

    470. 

  470. 		} else {							# assign new value to all 'open' descendants 

    471. 

  471. 			new.vv[match(all.D[!(all.D%in%remain.unchanged)],names(new.vv))] = nr.desc

    472. 

  472. 		}

    473. 

  473. 		new.vv[match(s, names(new.vv))]=nr.desc

    474. 

  474. 		lnHastingsRatio = log((K+1)/(2*N-2-K)) ### from Drummond and Suchard 2010: where N is tips, K is number of local parms in tree

    475. 

  475. 		lnPriorRatio = log(ptpois(K+1,lambda,nn)/ptpois(K,lambda,nn))

    476. 

  476. 		

    477. 

  477. 	} else {							# drop transition: MERGE

    478. 

  478. #		print("m")

    479. 

  479. 		decision="merge"

    480. 

  480. 		anc = get.ancestor.of.node(s, phy)

    481. 

  481. 		if(!is.root(anc, phy)) {			# base new rate on ancestral rate of selected branch

    482. 

  482. 			anc.vv=as.numeric(vv[match(anc,names(vv))])

    483. 

  483. 			na.vv=length(which(vv==anc.vv))

    484. 

  484. 			nr=(anc.vv*na.vv+cur.vv*ca.vv)/(ca.vv+na.vv)

    485. 

  485. 			new.vv[vv==cur.vv | vv==anc.vv]=nr

    486. 

  486. 		} else {							# if ancestor of selected node is root, base new rate on sister node

    487. 

  487. 			sister.tmp=get.desc.of.node(anc,phy)

    488. 

  488. 			sister=sister.tmp[sister.tmp!=s]

    489. 

  489. 			sis.vv=as.numeric(vv[match(sister,names(vv))])

    490. 

  490. 			ns.vv=length(which(vv==sis.vv))

    491. 

  491. 			nr=(sis.vv*ns.vv+cur.vv*ca.vv)/(ca.vv+ns.vv)

    492. 

  492. 			new.vv[vv==cur.vv | vv==sis.vv]=nr			

    493. 

  493. 		}

    494. 

  494. 		lnHastingsRatio = log((2*N-2-K+1)/K) ### from Drummond and Suchard 2010: where N is tips, K is number of local parms in tree

    495. 

  495. 		lnPriorRatio = log(ptpois(K-1,lambda,nn)/ptpois(K,lambda,nn))

    496. 

  496. 		

    497. 

  497. 	}

    498. 

  498. 	

    499. 

  499. 	return(list(new.delta=new.bb, new.values=new.vv, lnHastingsRatio=lnHastingsRatio, lnPriorRatio=lnPriorRatio, decision=decision))

    500. 

  500. }

    501. 

  501. 

    502. 

  502. switcheroo <- function(cur.values) { 

    503. 

  503. # mod 11.17.2010 JM Eastman

    504. 

  504. #	cur.delta=cur.delta.rates; cur.values=cur.rates; theta=FALSE; lambda=log(2); jumpsize=2

    505. 

  505. 

    506. 

  506. 	vv=cur.values

    507. 

  507. 	rates.sample=unique(vv)[sample(1:length(unique(vv)), 2)]

    508. 

  508. 	new.vv=vv

    509. 

  509. 	new.vv[which(vv==rates.sample[1])]=rates.sample[2]

    510. 

  510. 	new.vv[which(vv==rates.sample[2])]=rates.sample[1]

    511. 

  511. 	

    512. 

  512. 	return(list(new.values=new.vv))

    513. 

  513. }

    514. 

  514. 

    515. 

  515. generate.starting.point <- function(data, phy, node.des, theta=FALSE, K=FALSE, jumpsize) { 

    516. 

  516. # updated JME 11.14.2010

    517. 

  517. 	nn=length(phy$edge.length)

    518. 

  518. 	ntip=Ntip(phy)

    519. 

  519. 	if(!K) nshifts=rtpois(1,log(ntip),nn) else nshifts=K-1

    520. 

  520. 	if(nshifts!=0) bb=sample(c(rep(1,nshifts),rep(0,nn-nshifts)),replace=FALSE) else bb=rep(0,nn)

    521. 

  521. 	names(bb)=phy$edge[,2]

    522. 

  522. 	shifts=as.numeric(names(bb)[bb==1])

    523. 

  523. 	if(theta) {

    524. 

  524. 		min.max=c(adjust.value(min(data), jumpsize), adjust.value(max(data), jumpsize))

    525. 

  525. 		values=runif(sum(bb)+1, min=min.max[min.max==min(min.max)], max=min.max[min.max==max(min.max)])

    526. 

  526. 	} else {

    527. 

  527. 		init.rate=fitContinuous(phy,data)[[1]]$beta

    528. 

  528. 		min.max=c(adjust.rate(init.rate, jumpsize), adjust.rate(init.rate, jumpsize))

    529. 

  529. 		values=runif(sum(bb)+1, min=min.max[min.max==min(min.max)], max=min.max[min.max==max(min.max)])

    530. 

  530. 	}

    531. 

  531. 	internal.shifts<-tip.shifts<-numeric(0)

    532. 

  532. 	internal.shifts=sort(shifts[shifts>ntip])

    533. 

  533. 	tip.shifts=shifts[shifts<=ntip]

    534. 

  534. 	

    535. 

  535. 	if(length(internal.shifts)==0 & length(tip.shifts)==0) {

    536. 

  536. 		vv=rep(values, nn)

    537. 

  537. 	} else {

    538. 

  538. 		vv=bb

    539. 

  539. 		vv[]=values[length(values)]

    540. 

  540. 		i=0

    541. 

  541. 		if(length(internal.shifts)!=0) {

    542. 

  542. 			for(i in 1:length(internal.shifts)) {

    543. 

  543. 				d=node.des[which(names(node.des)==internal.shifts[i])]

    544. 

  544. 				vv[match(c(internal.shifts[i], unlist(d)), names(vv))]=values[i]

    545. 

  545. 			}

    546. 

  546. 		}

    547. 

  547. 		if(length(tip.shifts)!=0) {

    548. 

  548. 			for(j in 1:length(tip.shifts)) {

    549. 

  549. 				vv[match(tip.shifts[j], names(vv))]=values[j+i]

    550. 

  550. 			}

    551. 

  551. 		}

    552. 

  552. 	}

    553. 

  553. 	return(list(delta=unname(bb), values=unname(vv)))

    554. 

  554. }

    555. 

  555. 

    556. 

  556. assign.root.theta <- function(cur.theta, phy) {

    557. 

  557. 	root=phy$edge[1,1]

    558. 

  558. 	desc=phy$edge[which(phy$edge[,1]==root),2]

    559. 

  559. 	root.theta=cur.theta[sample(which(phy$edge[,2]==desc),1)]

    560. 

  560. 	root.theta

    561. 

  561. }

    562. 

  562. 

    563. 

  563. design.vcv.phylo=function(phy, array=TRUE) {

    564. 

  564. 	n=Ntip(phy)

    565. 

  565. 	tips=lapply(phy$edge[,2], function(x) get.descendants.of.node(x,phy,tips=TRUE))

    566. 

  566. 	if(!array) {

    567. 

  567. 		dd=matrix(0,n,n)

    568. 

  568. 		design=lapply(as.numeric(phy$edge[,2]), function(x) {

    569. 

  569. 				  new=dd

    570. 

  570. 				  node=x

    571. 

  571. 				  marker=as.numeric(which(phy$edge[,2]==node))

    572. 

  572. 				  if(node<=n) tt=node else tt=tips[[marker]]

    573. 

  573. 				  for(i in length(tt):1) {

    574. 

  574. 					j=tt[1:i]

    575. 

  575. 					new[tt[i],j]<-new[j,tt[i]]<-1

    576. 

  576. 				  }	  

    577. 

  577. 				  return(new)

    578. 

  578. 				  

    579. 

  579. 		}

    580. 

  580. 		)

    581. 

  581. 	} else {

    582. 

  582. 		design=array(0,dim=c(n,n,length(phy$edge[,2])))

    583. 

  583. 		for(k in 1:length(phy$edge[,2])){

    584. 

  584. 			node=as.numeric(phy$edge[k,2])

    585. 

  585. 			marker=k

    586. 

  586. 			if(node<=n) tt=node else tt=tips[[marker]]

    587. 

  587. 			for(i in length(tt):1) {

    588. 

  588. 				j=tt[1:i]

    589. 

  589. 				design[tt[i],j,k]<-design[j,tt[i],k]<-1

    590. 

  590. 			}	  

    591. 

  591. 		}		 

    592. 

  592. 	} 

    593. 

  593. 	return(design) 

    594. 

  594. }

    595. 

  595. 

    596. 

  596. update.vcv=function(ape.tre, new.rates) {

    597. 

  597. 	n=ape.tre

    598. 

  598. 	n$edge.length=n$edge.length*new.rates

    599. 

  599. 	vv=vcv.phylo(n)

    600. 

  600. 	return(vv)

    601. 

  601. }

    602. 

  602. 

    603. 

  603. update.vcv.older=function(ape.tre, design.vcv, new.rates) {

    604. 

  604. 	n=Ntip(ape.tre)

    605. 

  605. 	new.vcv=matrix(0,n,n)

    606. 

  606. 	new.edges=ape.tre$edge.length*new.rates

    607. 

  607. 	for(i in 1:length(new.rates)) {

    608. 

  608. 		new.vcv=new.vcv+new.edges[i]*design.vcv[[i]]

    609. 

  609. 	}

    610. 

  610. 	return(new.vcv)

    611. 

  611. }

    612. 

  612. 

    613. 

  613. update.vcv.new=function(ape.tre, new.rates, scalar.list) {

    614. 

  614. #	phy=birthdeath.tree(1,0,taxa=200); d=design.vcv.phylo(phy); s=build.scalar.sets(d); Rprof(); system.time(update.vcv.new(phy, rep(1,length(phy$edge.length)), s)); print(summaryRprof()); system.time(vcv.phylo(phy))

    615. 

  615. 	n=Ntip(ape.tre)

    616. 

  616. 	e=new.rates*ape.tre$edge.length

    617. 

  617. 	ss=scalar.list*e

    618. 

  618. 	entries=apply(ss,3,sum)

    619. 

  619. 	return(matrix(entries,n,n))

    620. 

  620. }

    621. 

  621. 

    622. 

  622. update.vcv.newest=function(ape.tre, new.rates, scalar.list) {

    623. 

  623. #	phy=birthdeath.tree(1,0,taxa=200); d=design.vcv.phylo(phy); s=build.scalar.sets(d); Rprof(); system.time(update.vcv.new(phy, rep(1,length(phy$edge.length)), s)); print(summaryRprof()); system.time(vcv.phylo(phy))

    624. 

  624. 	n=Ntip(ape.tre)

    625. 

  625. 	e=new.rates*ape.tre$edge.length

    626. 

  626. 	ss=scalar.list$scalars*e

    627. 

  627. 	new.vcv=matrix(0,n,n)

    628. 

  628. 	dg=which(scalar.list$diags)

    629. 

  629. 	cv=which(scalar.list$attr)

    630. 

  630. 	new.vcv[dg]=apply(scalar.list$scalars[,,dg],2,sum)

    631. 

  631. 	new.vcv[cv]=apply(scalar.list$scalars[,,cv],2,sum)

    632. 

  632. 	return(new.vcv)

    633. 

  633. }

    634. 

  634. 

    635. 

  635. 

    636. 

  636. build.scalar.sets=function(design.vcv) {

    637. 

  637. 	if(!is.array(design.vcv)) stop("Cannot process design matrix: must be array")

    638. 

  638. 	d=design.vcv

    639. 

  639. 	t=dim(design.vcv)[1]

    640. 

  640. 	ss=array(dim=c(1,(2*t-2),(t^2)))

    641. 

  641. 	index=0

    642. 

  642. 	scalar.attributes=c()

    643. 

  643. 	diags=c()

    644. 

  644. 	for(i in 1:t) {

    645. 

  645. 		for(j in 1:t) {

    646. 

  646. 			index=index+1

    647. 

  647. 			ss[,,index]=(as.logical(d[i,j,]))

    648. 

  648. 			if(any(ss[,,index])) scalar.attributes[index]=TRUE else scalar.attributes[index]=FALSE

    649. 

  649. 			if(i==j) diags[index]=TRUE else diags[index]=FALSE

    650. 

  650. 		}

    651. 

  651. 	}

    652. 

  652. 	for(i in 1:index) {

    653. 

  653. 		

    654. 

  654. 	}

    655. 

  655. 	return(list(scalars=ss, attr=scalar.attributes, diags=diags))

    656. 

  656. }

    657. 

  657. 

    658. 

  658. get.descendants.of.node <- function(node, phy, tips=FALSE) {

    659. 

  659. 	storage=list()

    660. 

  660. 	if(is.null(node)) node=phy$edge[1,1]

    661. 

  661. 	if(node%in%phy$edge[,1]) {

    662. 

  662. 		desc=c(sapply(node, function(x) {return(phy$edge[which(phy$edge[,1]==x),2])}))

    663. 

  663. 		while(any(desc%in%phy$edge[,1])) {

    664. 

  664. 			desc.true=desc[which(desc%in%phy$edge[,1])]

    665. 

  665. 			desc.desc=lapply(desc.true, function(x) return(get.desc.of.node(x, phy)))

    666. 

  666. 			

    667. 

  667. 			storage=list(storage,union(desc,desc.desc))

    668. 

  668. 			

    669. 

  669. 			desc=unlist(desc.desc)

    670. 

  670. 		}

    671. 

  671. 		storage=as.numeric(sort(unique(unlist(union(desc,storage)))))

    672. 

  672. 	}

    673. 

  673. 	if(tips) return(storage[storage<=Ntip(phy)]) else return(storage)

    674. 

  674. }

    675. 

  675. 

    676. 

  676. get.desc.of.node <- function(node, phy) {

    677. 

  677. 	return(phy$edge[which(phy$edge[,1]==node),2])

    678. 

  678. }

    679. 

  679. 

    680. 

  680. get.ancestor.of.node<-function(node, phy) {

    681. 

  681. 	anc=phy$edge[which(phy$edge[,2]==node),1]

    682. 

  682. 	anc

    683. 

  683. }

    684. 

  684. 

    685. 

  685. get.ancestors.of.node<-function(node, phy) {

    686. 

  686. 	ancestors=c()

    687. 

  687. 	a=0

    688. 

  688. 	while(1) {

    689. 

  689. 		a=a+1

    690. 

  690. 		node<-as.numeric(get.ancestor.of.node(node, phy))

    691. 

  691. 		if(!length(node)) break() else ancestors[a]=node

    692. 

  692. 	}

    693. 

  693. 	return(ancestors)

    694. 

  694. }

    695. 

  695. 

    696. 

  696. choose.one <- function(cur.delta)

    697. 

  697. {

    698. 

  698. 	bb=cur.delta

    699. 

  699. 	s=sample(1:length(bb),1)

    700. 

  700. 	bb[s]=1-bb[s]

    701. 

  701. 	bb

    702. 

  702. }

    703. 

  703. 

    704. 

  704. is.root<-function(node,phy) {

    705. 

  705. 	if(node==phy$edge[1,1]) return(TRUE) else return(FALSE)

    706. 

  706. }

    707. 

  707. 			   

    708. 

  708. within.range=function(x, min, max) {			 

    709. 

  709. 	a=sign(x-min)

    710. 

  710. 	b=sign(x-max)

    711. 

  711. 	if(abs(a+b)==2) return(FALSE) else return(TRUE)

    712. 

  712. }

    713. 

  713. 			   

    714. 

  714. 

    715. 

  715. assess.lnR <- function(lnR) {

    716. 

  716. 	if(is.na(lnR) || abs(lnR)==Inf) {

    717. 

  717. 		error=TRUE

    718. 

  718. 		r=0

    719. 

  719. 	} else {

    720. 

  720. 		if(lnR < -20) {

    721. 

  721. 			r=0 

    722. 

  722. 		} else if(lnR > 0) {

    723. 

  723. 			r=1 

    724. 

  724. 		} else {

    725. 

  725. 			r=exp(lnR)

    726. 

  726. 		}

    727. 

  727. 		error=FALSE

    728. 

  728. 	}

    729. 

  729. 	return(list(r=r, error=error))

    730. 

  730. }

    731. 

  731. 

    732. 

  732. apply.BMM.to.apetree<-function(apetree, rates) {

    733. 

  733. 	apetree$edge.length=apetree$edge.length*rates

    734. 

  734. 	return(apetree)

    735. 

  735. }

    736. 

  736. 

    737. 

  737. ape.to.ouch.tre<-function (tree, scale = FALSE, branch.lengths = tree$edge.length) 

    738. 

  738. {

    739. 

  739.     if (!inherits(tree, "phylo")) 

    740. 

  740. 	stop(sQuote("tree"), " must be of class ", sQuote("phylo"))

    741. 

  741.     nnodes <- nrow(tree$edge) + 1

    742. 

  742.     n.term <- length(tree$tip.label)

    743. 

  743.     n.int <- nnodes - n.term

    744. 

  744.     tmp <- matrix(NA, nnodes, 2)

    745. 

  745.     tmp[-1, 1:2] <- tree$edge

    746. 

  746.     tmp[1, 2] <- nnodes + 1

    747. 

  747.     bl <- c(0, branch.lengths)

    748. 

  748. 	bl.ape=bl

    749. 

  749. 	names(bl.ape)=c("root", tree$edge[,2])

    750. 

  750.     reord <- order(-tmp[, 2])

    751. 

  751.     bl <- bl[reord]

    752. 

  752. 	bl.ouch <- bl.ape[reord]

    753. 

  753.     tmp <- tmp[reord, ]

    754. 

  754.     node <- seq(nnodes)

    755. 

  755.     ancestor <- rep(NA, nnodes)

    756. 

  756.     for (n in 2:nnodes) {

    757. 

  757.         anc <- which(tmp[, 2] == tmp[n, 1])

    758. 

  758.         if (length(anc) > 1) 

    759. 

  759. 		stop("invalid tree")

    760. 

  760.         if (length(anc) > 0) {

    761. 

  761.             ancestor[n] <- node[anc]

    762. 

  762.         }

    763. 

  763.         else {

    764. 

  764.             ancestor[n] <- node[1]

    765. 

  765.         }

    766. 

  766.     }

    767. 

  767.     if (is.null(tree$node.label)) 

    768. 

  768. 	tree$node.label <- rep("", n.int)

    769. 

  769.     species <- rev(c(tree$tip.label, tree$node.label[-1], tree$node.label[1]))

    770. 

  770.     times <- rep(NA, nnodes)

    771. 

  771.     for (n in 1:nnodes) times[n] <- ouch:::branch.height(node, ancestor, bl, n)

    772. 

  772.     if (is.logical(scale)) {

    773. 

  773.         if (is.na(scale)) 

    774. 

  774. 		stop("if ", sQuote("scale"), " is logical, it must be either true or false")

    775. 

  775.         if (scale) 

    776. 

  776. 		times <- times/max(times)

    777. 

  777.     }

    778. 

  778.     else if (is.numeric(scale)) {

    779. 

  779.         times <- times/abs(scale)

    780. 

  780.     }

    781. 

  781.     else {

    782. 

  782.         stop(sQuote("scale"), " must be either logical or numeric")

    783. 

  783.     }

    784. 

  784.     return(list(ouch.tre=ouchtree(nodes = node, ancestors = ancestor, times = times, labels = species), ape.ordered=bl.ouch))

    785. 

  785. }

    786. 

  786. 

    787. 

  787. adjust.value <- function(value, jumpsize, theta=FALSE) {

    788. 

  788. # mod 10.20.2010 JM Eastman

    789. 

  789. 	vv=value

    790. 

  790. 	if(runif(1)<0.5 | length(unique(vv))==1) {

    791. 

  791. 		rch <- runif(1, min = -jumpsize, max = jumpsize)

    792. 

  792. 		return(vv[]+rch)

    793. 

  793. 	} else {

    794. 

  794. 		return((vv-mean(vv))*runif(1,min=-jumpsize, max=jumpsize)+mean(vv))

    795. 

  795. 	}

    796. 

  796. }

    797. 

  797. 

    798. 

  798. adjust.rate <- function(rate, jumpsize) {

    799. 

  799. # mod 10.20.2010 JM Eastman

    800. 

  800. 	vv=rate

    801. 

  801. 	v=log(vv)

    802. 

  802. 	rch <- runif(1, min = -jumpsize, max = jumpsize)

    803. 

  803. 	v=exp(v[]+rch)

    804. 

  804. 	v

    805. 

  805. }

    806. 

  806. 

    807. 

  807. prepare.data <- function(phy, data) {

    808. 

  808. 	td <- treedata(phy, data, sort = T)	

    809. 

  809. 	return(list(ape.tre=td$phy, orig.dat=td$data[,1]))

    810. 

  810. }

    811. 

  811. 

    812. 

  812. prepare.ouchdata <- function(phy, data) {

    813. 

  813. 	td <- treedata(phy, data, sort = T)	

    814. 

  814. 	

    815. 

  815. 	ot <- ape.to.ouch.tre(td$phy->ape.tre, scale=FALSE)		

    816. 

  816. 	ouch.tre=ot$ouch.tre

    817. 

  817. 	ouch.dat=data.frame(as.numeric(ouch.tre@nodes), as.character(ouch.tre@nodelabels))

    818. 

  818. 	names(ouch.dat)=c("nodes","labels")

    819. 

  819. 	ouch.dat[ouch.dat[,2]=="",2]=NA

    820. 

  820. 	ouch.dat$trait=NA

    821. 

  821. 	mM=match(ouch.dat$labels,names(data))

    822. 

  822. 	for(i in 1:nrow(ouch.dat)){

    823. 

  823. 		if(!is.na(mM[i])){

    824. 

  824. 			ouch.dat$trait[i]=data[mM[i]]	

    825. 

  825. 		}

    826. 

  826. 	}

    827. 

  827. 	return(list(ouch.tre=ouch.tre, ouch.dat=ouch.dat, ape.tre=td$phy, orig.dat=td$data[,1], ape.ordered=ot$ape.ordered))

    828. 

  828. }

    829. 

  829. 

    830. 

  830. summarize.run<-function(cOK, endparms, cur.model) {

    831. 

  831. #	endparmsOU = c(nRateProp, nRootProp, nAlphaProp, nTcatProp, nThetaProp)

    832. 

  832. 

    833. 

  833. 	if(cur.model=="BM") {

    834. 

  834. 		names(endparms)<-names(cOK)<-c("rates", "rate.swap", "rate.catg", "root")

    835. 

  835. 	} else {

    836. 

  836. 		names(endparms)<-names(cOK)<-c("rate", "root", "alpha", "regimes", "theta")

    837. 

  837. 	}

    838. 

  838. 	names(endparms)=paste("pr.",names(endparms),sep="")

    839. 

  839. 	names(cOK)=paste("ok.",names(cOK),sep="")

    840. 

  840. 	cat("\n")

    841. 

  841. 	if(cur.model=="BM") {

    842. 

  842. 		print(endparms[paste("pr.", c("rates", "rate.swap", "rate.catg", "root"),sep="")])

    843. 

  843. 		print(cOK[paste("ok.", c("rates", "rate.swap", "rate.catg", "root"),sep="")])

    844. 

  844. 	} else {

    845. 

  845. 		print(endparms[paste("pr.", c("rate", "root", "alpha", "regimes", "theta"),sep="")])

    846. 

  846. 		print(cOK[paste("ok.", c("rate", "root", "alpha", "regimes", "theta"),sep="")])

    847. 

  847. 	}

    848. 

  848. }

    849. 

  849. 

    850. 

  850. generate.log<-function(bundled.parms, cur.model, file, init=FALSE) {

    851. 

  851. #	bundled.parms=list(gen=i, mrate=mean(cur.rates), cats=sum(cur.delta.rates)+1, root=cur.root, alpha=cur.alpha, reg=sum(cur.delta.theta)+1, theta=mean(cur.theta), lnL=curr.lnL)

    852. 

  852. 	res=bundled.parms

    853. 

  853. 

    854. 

  854. 	if(init) {

    855. 

  855. 		if(cur.model=="BM") msg=paste("gen", "model", "mean.rate", "rates", "root", "lnL", sep="\t") else msg=paste("gen", "model", "rate", "root", "alpha", "regimes", "mean.theta", "lnL", sep="\t")

    856. 

  856. 	} else {

    857. 

  857. 		if(cur.model=="BM") {

    858. 

  858. 			msg<-paste(res$gen, sQuote(cur.model), sprintf("%.3f", res$mrate), res$cats, sprintf("%.3f", res$root), sprintf("%.3f", res$lnL),sep="\t")

    859. 

  859. 		} else {

    860. 

  860. 			msg<-paste(res$gen, sQuote(cur.model), sprintf("%.3f", res$rate), sprintf("%.2f", res$root), sprintf("%.4f", res$alpha), res$regimes, sprintf("%.3f", res$mean.theta), sprintf("%.3f", res$lnL),sep="\t")

    861. 

  861. 		}

    862. 

  862. 	}

    863. 

  863. 	write(msg, file=file, append=TRUE)

    864. 

  864. }

    865. 

  865. 

    866. 

  866. parlogger<-function(model, init=FALSE, node.des, i, curr.lnL, cur.root=NULL, cur.rates=NULL, cur.delta.rates=NULL, cur.alpha=NULL, cur.theta=NULL, cur.delta.theta=NULL, parmBase) {	

    867. 

  867. 	if(init) {

    868. 

  868. 		msg.long=names(node.des[-1])

    869. 

  869. 		if(model=="BM") {

    870. 

  870. 			msg.short=paste("gen", "root", "lnL", sep="\t") 

    871. 

  871. 			parlogs=paste(parmBase, paste(c("summary", "rates", "rate.shifts"), "txt", sep="."), sep="")

    872. 

  872. 			sapply(parlogs[1], function(x) write.table(msg.short, x, quote=FALSE, col.names=FALSE, row.names=FALSE))

    873. 

  873. 			sapply(parlogs[2:length(parlogs)], function(x) write.table(paste(msg.long,collapse=" "), x, quote=FALSE, col.names=FALSE, row.names=FALSE))

    874. 

  874. 		} else {

    875. 

  875. 			msg.short=paste("gen", "root", "rate", "alpha", "lnL", sep="\t")

    876. 

  876. 			parlogs=paste(parmBase, paste(c("summary", "optima", "optima.shifts"), "txt", sep="."), sep="")

    877. 

  877. 			sapply(parlogs[1], function(x) write.table(msg.short, x, quote=FALSE, col.names=FALSE, row.names=FALSE))

    878. 

  878. 			sapply(parlogs[2:length(parlogs)], function(x) write.table(paste(msg.long,collapse=" "), x, quote=FALSE, col.names=FALSE, row.names=FALSE))

    879. 

  879. 		}

    880. 

  880. 	} else {

    881. 

  881. 		if(model=="BM") {

    882. 

  882. 			parlogs=paste(parmBase, paste(c("summary", "rates", "rate.shifts"), "txt", sep="."), sep="")

    883. 

  883. 			outlist=list(paste(i, cur.root, curr.lnL, sep="\t"), cur.rates, cur.delta.rates) 

    884. 

  884. 			sapply(1:length(outlist), function(x) write.table(paste(outlist[[x]],collapse=" "), parlogs[[x]], quote=FALSE, col.names=FALSE, row.names=FALSE, append=TRUE)) 

    885. 

  885. 		} else {

    886. 

  886. 			parlogs=paste(parmBase, paste(c("summary", "optima", "optima.shifts"), "txt", sep="."), sep="")

    887. 

  887. 			outlist=list(paste(i, cur.root, cur.rates, cur.alpha, curr.lnL, sep="\t"), cur.theta, cur.delta.theta) 

    888. 

  888. 			sapply(1:length(outlist), function(x) write.table(paste(outlist[[x]],collapse=" "), parlogs[[x]], quote=FALSE, col.names=FALSE, row.names=FALSE, append=TRUE)) 

    889. 

  889. 		}

    890. 

  890. 	}

    891. 

  891. }

    892. 

  892. 

    893. 

  893. generate.error.message<-function(i, mod.cur, mod.new, lnR, errorLog, init=FALSE) {

    894. 

  894. 	if(init) {

    895. 

  895. 		initmsg = write(paste("gen", "curr.lnL", "new.lnL", "lnR", sep="\t"), file=errorLog)

    896. 

  896. 	} else {

    897. 

  897. 		write(paste(i, sprintf("%.3f", mod.cur$lnL), sprintf("%.3f", mod.new$lnL), sprintf("%.3f", lnR), sep="\t"), file=errorLog)

    898. 

  898. 	}

    899. 

  899. }

    900. 

  900. 

    901. 

  901. determine.accepted.proposal<-function(startparms, endparms, cOK) {

    902. 

  902. 	which(startparms!=endparms)->marker

    903. 

  903. 	cOK[marker]=cOK[marker]+1

    904. 

  904. 	cOK

    905. 

  905. }

    906. 

  906. 

    907. 

  907. rtpois<-function(N, lambda, k) {

    908. 

  908. 	p=ppois(k, lambda, lower.tail=FALSE)

    909. 

  909. 	out=qpois(runif(N, min=0, max=1-p), lambda)

    910. 

  910. 	out

    911. 

  911. }

    912. 

  912. 				 

    913. 

  913. ptpois<-function(x, lambda, k) {

    914. 

  914. 	p.k=ppois(k, lambda, lower.tail=FALSE)

    915. 

  915. 	p.x=ppois(x, lambda, lower.tail=FALSE)

    916. 

  916. 	ptp=p.x/(1-p.k)

    917. 

  917. 	return(ptp)

    918. 

  918. }

    919. 

  919. 				 

    920. 

  920. ## SUMMARIZING MCMC -- PLOTTING FUNCTIONS ##

    921. 

  921. shifts.plot=function(phy, base.dir, burnin=0, level=0.03, internal.only=FALSE, paint.branches=TRUE, pdf=TRUE, verbose=TRUE, ...) {

    922. 

  922. #	phy; base.dir; burnin=0; level=0.03; internal.only=FALSE; paint.branches=TRUE; pdf=TRUE; verbose=TRUE

    923. 

  923. 	nullphy=phy

    924. 

  924. 	nullphy$tip.label[]=""

    925. 

  925. 	color.length=21

    926. 

  926. 	require(ape)

    927. 

  927. 	oldwd=getwd()

    928. 

  928. 	setwd(base.dir)

    929. 

  929. 	files=dir()

    930. 

  930. 	if(any(grep("optima", files))) optima=TRUE else optima=FALSE

    931. 

  931. 	if(optima) {

    932. 

  932. 		shifts.file=files[grep("optima.shifts.txt", files)]

    933. 

  933. 		estimates.file=files[grep("optima.txt", files)]

    934. 

  934. 	} else {

    935. 

  935. 		shifts.file=files[grep("rate.shifts.txt", files)]

    936. 

  936. 		estimates.file=files[grep("rates.txt", files)]

    937. 

  937. 	}

    938. 

  938. 	if(pdf | verbose) {

    939. 

  939. 		lab=paste(strsplit(base.dir, ".", fixed=TRUE)[[1]][1:2],collapse=".")

    940. 

  940. 		out.base=paste(lab, paste("bi", burnin, sep=""), paste("fq", level, sep=""), sep=".")

    941. 

  941. 		outdir=paste("./results")

    942. 

  942. 		if(!file.exists(outdir)) dir.create(outdir)

    943. 

  943. 	}

    944. 

  944. 	cat("READING shifts...\n")

    945. 

  945. 	results=read.table(shifts.file)

    946. 

  946. 	names(results)=as.numeric(results[1,])

    947. 

  947. 	burnin=ceiling(burnin*nrow(results))

    948. 

  948. 	results=results[-c(1:(burnin+1)),]

    949. 

  949. 	if(!all(names(results)%in%phy$edge[,2])) stop("Cannot process results: check that tree matches posterior summaries")

    950. 

  950. 	hits.tmp<-shifts<-apply(results,1,sum)

    951. 

  951. 	hits=length(hits.tmp[hits.tmp>0])

    952. 

  952. 	aa.tmp=apply(results, 2, function(x) sum(x))

    953. 

  953. 	aa.tmp=aa.tmp/hits

    954. 

  954. 	aa=aa.tmp[aa.tmp>=level]

    955. 

  955. 	aa=aa[order(aa, decreasing=TRUE)]

    956. 

  956. 	aa.nodes=aa[which(as.numeric(names(aa))>Ntip(phy))]

    957. 

  957. 	aa.tips=aa[which(as.numeric(names(aa))<=Ntip(phy))]

    958. 

  958. 	aa=aa.nodes

    959. 

  959. 	nodes=as.numeric(names(aa))

    960. 

  960. 	nodes=nodes[nodes>Ntip(phy)]

    961. 

  961. 	desc=lapply(nodes, function(x) {

    962. 

  962. 				foo=get.descendants.of.node(x,phy)

    963. 

  963. 				if(length(foo)) return(phy$tip.label[foo[foo<=Ntip(phy)]]) else return(NULL)

    964. 

  964. 				}

    965. 

  965. 				)

    966. 

  966. 	cat("READING estimates...\n")

    967. 

  967. 	ests=read.table(estimates.file)

    968. 

  968. 	names(ests)=ests[1,]

    969. 

  969. 	ests=ests[-c(1:(burnin+1)),]

    970. 

  970. 	average.ests.tmp=apply(ests,2,mean)

    971. 

  971. 	average.ests=average.ests.tmp-median(average.ests.tmp)

    972. 

  972. 	if(paint.branches) {

    973. 

  973. 		require(colorspace)

    974. 

  974. 		cce=diverge_hcl(2*color.length, power = 0.5)

    975. 

  975. 		e.seq=seq(min(average.ests),max(average.ests),length=color.length)

    976. 

  976. 		color.gamut=seq(-max(abs(e.seq)), max(abs(e.seq)), length=length(cce))

    977. 

  977. 		colors.branches=sapply(average.ests, function(x) {y=cce[which(min(abs(color.gamut-x))==abs(color.gamut-x))]; return(y[sample(1:length(y),1)])})

    978. 

  978. 		colors.branches=colors.branches[match(as.numeric(names(colors.branches)), phy$edge[,2])]

    979. 

  979. 	} else {

    980. 

  980. 		colors.branches=1

    981. 

  981. 	}

    982. 

  982. 	ccx=diverge_hcl(color.length, c = c(100, 0), l = c(50, 90), power = 1.1)

    983. 

  983. 	c.seq=round(seq(-1,1,by=0.1),digits=1)

    984. 

  984. 	

    985. 

  985. 	if(length(nodes)) {

    986. 

  986. 		require(colorspace)

    987. 

  987. 		cols=sapply(nodes, function(x) {

    988. 

  988. 					a=get.ancestor.of.node(x, phy)

    989. 

  989. 					comp=ests[,which(names(ests)==a)]

    990. 

  990. 					this=ests[,which(names(ests)==x)]

    991. 

  991. 					if(!length(comp)) { # dealing with first descendant of root

    992. 

  992. 					d=get.desc.of.node(a, phy)

    993. 

  993. 					d=d[which(d!=x)]

    994. 

  994. 					

    995. 

  995. # find if sister descendant also experienced rate shift 

    996. 

  996. 					d.shifts=results[, which(names(results)==d)]

    997. 

  997. 					comp=ests[,which(names(ests)==d)]

    998. 

  998. 					x.res=sapply(1:length(d.shifts), function(y) {

    999. 

  999. 								 if(d.shifts[y]==0) {

    1000. 

  1000. 								 zz=this[y]-comp[y]

    1001. 

  1001. 								 if(zz>0) {

    1002. 

  1002. 								 return(1) 

    1003. 

  1003. 								 } else {

    1004. 

  1004. 								 if(zz<0) return(-1) else return(0)

    1005. 

  1005. 								 }

    1006. 

  1006. 								 } else {

    1007. 

  1007. 								 return(0)

    1008. 

  1008. 								 }

    1009. 

  1009. 								 }

    1010. 

  1010. 								 )

    1011. 

  1011. 					x.res=mean(x.res[x.res!=0])

    1012. 

  1012. 					if(is.na(x.res)) x.res=0

    1013. 

  1013. 					} else {

    1014. 

  1014. 					yy=this-comp

    1015. 

  1015. 					zz=yy

    1016. 

  1016. 					zz[yy>0]=1

    1017. 

  1017. 					zz[yy<0]=-1

    1018. 

  1018. 					zz[yy==0]=0

    1019. 

  1019. 					x.res=mean(zz[zz!=0])

    1020. 

  1020. 					if(is.na(x.res)) x.res=0

    1021. 

  1021. 					}

    1022. 

  1022. 					return(x.res)

    1023. 

  1023. 					}

    1024. 

  1024. 					)

    1025. 

  1025. 		colors.nodes=ccx[match(round(cols,1),c.seq)]

    1026. 

  1026. 		names(colors.nodes)=nodes

    1027. 

  1027. 		colors.nodes=colors.nodes[which(as.numeric(names(colors.nodes))>Ntip(phy))]

    1028. 

  1028. 	} else {

    1029. 

  1029. 		colors.nodes=NULL

    1030. 

  1030. 		cols=NULL

    1031. 

  1031. 	}

    1032. 

  1032. 	

    1033. 

  1033. 	if(length(aa.tips) & !internal.only) {

    1034. 

  1034. 		tt<-tt.cex<-rep(0,Ntip(phy))

    1035. 

  1035. 		tt[as.numeric(names(aa.tips))]=1

    1036. 

  1036. #		tt.cex[as.numeric(names(aa.tips))]=aa.tips/max(aa.tmp)

    1037. 

  1037. 		tt.cex[as.numeric(names(aa.tips))]=aa.tips

    1038. 

  1038. 		

    1039. 

  1039. 		tt.col=sapply(names(aa.tips), function(x) {

    1040. 

  1040. 					  a=get.ancestor.of.node(x, phy)

    1041. 

  1041. 					  comp=ests[,which(names(ests)==a)]

    1042. 

  1042. 					  this=ests[,which(names(ests)==x)]

    1043. 

  1043. 					  if(!length(comp)) { # dealing with first descendant of root

    1044. 

  1044. 					  d=get.desc.of.node(a, phy)

    1045. 

  1045. 					  d=d[which(d!=x)]

    1046. 

  1046. 					  

    1047. 

  1047. # find if sister descendant also experienced rate shift 

    1048. 

  1048. 					  d.shifts=results[, which(names(results)==d)]

    1049. 

  1049. 					  comp=ests[,which(names(ests)==d)]

    1050. 

  1050. 					  x.res=sapply(1:length(d.shifts), function(y) {

    1051. 

  1051. 								   if(d.shifts[y]==0) {

    1052. 

  1052. 								   zz=this[y]-comp[y]

    1053. 

  1053. 								   if(zz>0) {

    1054. 

  1054. 								   return(1) 

    1055. 

  1055. 								   } else {

    1056. 

  1056. 								   if(zz<0) return(-1) else return(0)

    1057. 

  1057. 								   }

    1058. 

  1058. 								   } else {

    1059. 

  1059. 								   return(0)

    1060. 

  1060. 								   }

    1061. 

  1061. 								   }

    1062. 

  1062. 								   )

    1063. 

  1063. 					  x.res=mean(x.res[x.res!=0])

    1064. 

  1064. 					  } else {

    1065. 

  1065. 					  yy=this-comp

    1066. 

  1066. 					  zz=yy

    1067. 

  1067. 					  zz[yy>0]=1

    1068. 

  1068. 					  zz[yy<0]=-1

    1069. 

  1069. 					  zz[yy==0]=0

    1070. 

  1070. 					  x.res=mean(zz[zz!=0])

    1071. 

  1071. 					  }

    1072. 

  1072. 					  return(x.res)

    1073. 

  1073. 					  }

    1074. 

  1074. 					  )

    1075. 

  1075. 		tt.ccx=diverge_hcl(21, c = c(100, 0), l = c(50, 90), power = 1.1)

    1076. 

  1076. 		tt.c.seq=round(seq(-1,1,by=0.1),digits=1)

    1077. 

  1077. 		tt.colors.nodes=tt.ccx[match(round(tt.col,1),tt.c.seq)]

    1078. 

  1078. 		names(tt.colors.nodes)=as.numeric(names(aa.tips))

    1079. 

  1079. 		colors.tips.tmp=tt.colors.nodes[order(as.numeric(names(tt.colors.nodes)))]

    1080. 

  1080. 		colors.tips=rep(NA, Ntip(phy))

    1081. 

  1081. 		colors.tips[as.numeric(names(colors.tips.tmp))]=colors.tips.tmp

    1082. 

  1082. 		tip.names=as.list(phy$tip.label[as.numeric(names(aa.tips))])

    1083. 

  1083. 		names(tip.names)=paste("node",names(aa.tips),sep=".")

    1084. 

  1084. 	}	else {

    1085. 

  1085. 		tt.cex=NULL

    1086. 

  1086. 		tt.col=NULL

    1087. 

  1087. 		tip.names=NULL

    1088. 

  1088. 		colors.tips=NULL

    1089. 

  1089. 		tt=rep(0,Ntip(phy))

    1090. 

  1090. 	}

    1091. 

  1091. 	

    1092. 

  1092. ## PLOTTING OF TREE

    1093. 

  1093. 	CEX=max(c(0.05, (2+-0.36*log(Ntip(phy))) ))	

    1094. 

  1094. 	if(pdf) pdf(file=paste(outdir, paste(out.base,"pdf",sep="."),sep="/"))

    1095. 

  1095. 	plot(phy, cex=CEX, lwd=0.4, edge.width=0.5, edge.color=colors.branches, show.tip.label=TRUE, label.offset=strwidth(par("pch"),cex=1.25), ...)

    1096. 

  1096. 	nn=(Ntip(phy)+1):max(phy$edge[,2])

    1097. 

  1097. 	ll<-cc<-rr<-rep(0,length(nn))

    1098. 

  1098. 	ll[nn%in%nodes]=1

    1099. 

  1099. #	if(length(nodes)) cc[nn%in%nodes]=aa[match(nn[nn%in%nodes],nodes)]/max(c(aa,tt.cex)) 

    1100. 

  1100. 	if(length(nodes)) cc[nn%in%nodes]=aa[match(nn[nn%in%nodes],nodes)]

    1101. 

  1101. 	rr[nn%in%nodes]=colors.nodes[order(names(colors.nodes))]

    1102. 

  1102. 	nodelabels(pch=ifelse(ll==1, 21, NA), cex=2*cc, col=rr, bg=rr, lwd=0.5)

    1103. 

  1103. 	

    1104. 

  1104. 	tiplabels(pch=ifelse(tt==1, 21, NA), cex=2*tt.cex, col=colors.tips, bg=colors.tips, lwd=0.5)

    1105. 

  1105. 	legend("topright", title=toupper("direction"), cex=0.5, pt.cex=1, text.col="darkgray", legend = sprintf("%6.1f", rev(c.seq[seq(1,color.length,by=2)])), pch=21, ncol=1, col = "darkgray", pt.bg = rev(ccx[seq(1,color.length,by=2)]), box.lty="blank", border="white")

    1106. 

  1106. 	if(length(nodes) | (length(aa.tips) & !internal.only)) {

    1107. 

  1107. 		legend("bottomright", title=toupper("frequency"), cex=0.5, pt.cex=2*(seq(0, 1, length=9)), text.col="darkgray", legend = sprintf("%10.3f", seq(0, 1, length=9)), pch=21, ncol=1, col = "darkgray", pt.bg = "white", box.lty="blank", border="white")

    1108. 

  1108. 	}

    1109. 

  1109. 	if(pdf) dev.off()

    1110. 

  1110. 	

    1111. 

  1111. 	if(length(nodes)) {

    1112. 

  1112. 		names(cols)=names(aa)

    1113. 

  1113. 		names(desc)=paste("node",nodes,sep=".")

    1114. 

  1114. 	}

    1115. 

  1115. ##

    1116. 

  1116. 	

    1117. 

  1117. 	if(verbose) {

    1118. 

  1118. 		summary.table=c(hits/nrow(results), mean(shifts), median(shifts))

    1119. 

  1119. 		names(summary.table)=c("shift.prob", "mean.shift","median.shifts")

    1120. 

  1120. 		write.table(summary.table, file=paste(outdir, paste(out.base, "run.summary.txt", sep="."), sep="/"), quote=FALSE, col.names=FALSE)

    1121. 

  1121. 		if(length(nodes)) {

    1122. 

  1122. 			for(nn in 1:length(nodes)) {

    1123. 

  1123. 				write.table(c(nodes[nn], desc[[nn]]), file=paste(outdir, paste(out.base, "shift.descendants.txt", sep="."), sep="/"), append=TRUE, quote=FALSE, row.names=FALSE, col.names=FALSE)

    1124. 

  1124. 			}

    1125. 

  1125. 		}

    1126. 

  1126. 		if(length(aa.tips)) {

    1127. 

  1127. 			for(nn in 1:length(aa.tips)) {

    1128. 

  1128. 				write.table(c(as.numeric(names(aa.tips)[nn]), tip.names[[nn]]), file=paste(outdir, paste(out.base, "shift.descendants.txt", sep="."), sep="/"), append=TRUE, quote=FALSE, row.names=FALSE, col.names=FALSE)

    1129. 

  1129. 			}

    1130. 

  1130. 		}

    1131. 

  1131. 		allres=data.frame(matrix(NA, nrow=nrow(phy$edge), ncol=3))

    1132. 

  1132. 		if(length(aa)) {

    1133. 

  1133. 			x=as.numeric(names(aa))

    1134. 

  1134. 			nodres=merge(as.data.frame(aa),as.data.frame(cols),by=0)

    1135. 

  1135. 			allres[nodres[,1],]=nodres

    1136. 

  1136. 		}

    1137. 

  1137. 		if(length(aa.tips)) {

    1138. 

  1138. 			x=as.numeric(names(aa.tips))

    1139. 

  1139. 			tipres=merge(as.data.frame(aa.tips), as.data.frame(tt.col),by=0)

    1140. 

  1140. 			allres[tipres[,1],]=tipres

    1141. 

  1141. 		}

    1142. 

  1142. 		allres=allres[!apply(allres, 1, function(x)any(is.na(x))),]

    1143. 

  1143. 		allres=allres[order(allres[,2], decreasing=TRUE),]

    1144. 

  1144. 		if(nrow(allres)>0) {

    1145. 

  1145. 			allres=data.frame(allres)

    1146. 

  1146. 			names(allres)=c("node", "conditional.shift.probability", "relative.shift.direction")

    1147. 

  1147. 			write.table(allres, file=paste(outdir, paste(out.base, "estimates.summary.txt", sep="."), sep="/"), quote=FALSE, row.names=FALSE)

    1148. 

  1148. 		}

    1149. 

  1149. 	}

    1150. 

  1150. 	setwd(oldwd)

    1151. 

  1151. 	

    1152. 

  1152. 	if(!optima) {

    1153. 

  1153. 		return(list(desc=desc, tips=tip.names, shift.prob=hits/nrow(results), mean.shifts=mean(shifts), median.shifts=median(shifts), nodeshift.prop=aa, nodeshift.direction=cols, tipshift.prop=aa.tips, tipshift.direction=tt.col, mean.rates=average.ests.tmp, mean.shifts=aa.tmp/nrow(results)))

    1154. 

  1154. 	} else {

    1155. 

  1155. 		return(list(desc=desc, tips=tip.names, shift.prob=hits/nrow(results), mean.shifts=mean(shifts), median.shifts=median(shifts), nodeshift.prop=aa, nodeshift.direction=cols, tipshift.prop=aa.tips, tipshift.direction=tt.col, mean.optima=average.ests.tmp, mean.shifts=aa.tmp/nrow(results)))

    1156. 

  1156. 	}

    1157. 

  1157. }

    1158. 

  1158. 

    1159. 

  1159. 

    1160. 

  1160. 

    1161. 

  1161. plot.quant<-function(phy, data, data.names, shrinker=1, relative=TRUE, cex=0.5, adj=c(15,5), lwd=0.5, font=1, ...){

    1162. 

  1162. 	std=function(x,max,min){return((x-min)/(max-min))}

    1163. 

  1163. 	if(relative) {

    1164. 

  1164. 		data=sapply(data, function(x) std(x, max(data), min(data)))

    1165. 

  1165. 	}

    1166. 

  1166. 	

    1167. 

  1167. 	f=match(phy$tip.label, data.names)

    1168. 

  1168. 	if(is.na(sum(f))) {

    1169. 

  1169. 		stop("tips cannot be matched to tree")

    1170. 

  1170. 	}

    1171. 

  1171. 	else {

    1172. 

  1172. 		data=data[f]

    1173. 

  1173. 		phy$trait=data

    1174. 

  1174. 		plot.phylo(phy, show.tip.label=T, label.offset=adj[1], cex=cex, font=font, ...)

    1175. 

  1175. 		tiplabels(text=NULL, pch=21, adj=adj[2], cex=c(phy$trait/shrinker), bg="white",lwd=lwd)

    1176. 

  1176. 		} 

    1177. 

  1177. }

    1178. 

  1178. 

    1179. 

  1179. branchcol.plot=function(phy, cur.rates, color.length=4, ...) {

    1180. 

  1180. 	color.length=color.length

    1181. 

  1181. 	require(ape)

    1182. 

  1182. 	require(colorspace)

    1183. 

  1183. 	ests=cur.rates

    1184. 

  1184. 	names(ests)=phy$edge[,2]

    1185. 

  1185. 	cce=diverge_hcl(color.length, power = 0.5)

    1186. 

  1186. 	e.seq=seq(min(ests),max(ests),length=color.length)

    1187. 

  1187. 	colors.branches=sapply(ests, function(x) cce[which(min(abs(e.seq-x))==abs(e.seq-x))])

    1188. 

  1188. 	names(colors.branches)=names(ests)

    1189. 

  1189. 	colors.branches=colors.branches[match(as.numeric(names(colors.branches)), phy$edge[,2])]	

    1190. 

  1190. 	plot(phy, cex=0.1, lwd=0.4, edge.width=0.5, edge.color=colors.branches, ...)

    1191. 

  1191. }

    1192. 

  1192. 

    1193. 

  1193. 

    1194. 

  1194. prior.posterior.plot=function(data, rpois.lambda) {

    1195. 

  1195. # data should be a vector of number of shifts, for instance, for all interations across the MCMC sampling (post-burnin)

    1196. 

  1196. 	require(ggplot2)

    1197. 

  1197. 	dataset <- data.frame(variable = gl(2, length(data), labels = c("posterior", "prior")), value = c(data, rpois(length(data),rpois.lambda))) 

    1198. 

  1198. 	ggplot(data = dataset, aes(x = value, fill = variable)) + geom_histogram(position = "dodge") + scale_fill_manual(values = c("gray", "black"))	

    1199. 

  1199. }

    1200. 

  1200. 

    1201. 

  1201. prior.posterior.plot.alt=function(outlogs, lambda, ...){

    1202. 

  1202. 	foo=outlogs

    1203. 

  1203. 	d=lapply(dir(foo),function(x) read.table(paste(foo,x,sep="/"),header=T))

    1204. 

  1204. 	names=dir(foo)

    1205. 

  1205. 	lapply(names,function(x)paste(strsplit(x,".",fixed=T)[[1]][1:2],collapse="."))->nn

    1206. 

  1206. 	

    1207. 

  1207. 	

    1208. 

  1208. 	n=length(nn)

    1209. 

  1209. 	ss=(1:10)^2

    1210. 

  1210. 	sn=ss-n

    1211. 

  1211. 	s=ss[which(sn==min(sn[sn>=0]))]

    1212. 

  1212. 	x=sqrt(s)

    1213. 

  1213. 	layout(matrix(1:s, x, x))

    1214. 

  1214. 		

    1215. 

  1215. 	for(i in 1:length(d)){

    1216. 

  1216. 		if(any(names(d[[i]])=="regimes")) {

    1217. 

  1217. 			plot(density(d[[i]]$regimes), bty="n",xlim=c(0,10), main=nn[[i]],xlab="regimes") 

    1218. 

  1218. 		} else {

    1219. 

  1219. 			plot(density(d[[i]]$rates),bty="n",xlim=c(0,10), main=nn[[i]],xlab="rates")

    1220. 

  1220. 		}

    1221. 

  1221. 		lines(density(rpois(nrow(d[[i]]),lambda)), ...)

    1222. 

  1222. 	}	

    1223. 

  1223. }

    1224. 

  1224. 

    1225. 

  1225. split.rate=function(cur.vv, n.desc, n.split, factor=log(2)){

    1226. 

  1226. 	dev=cur.vv-exp(log(cur.vv)+runif(1, -factor, factor))

    1227. 

  1227. 	nr.desc=cur.vv+dev/n.desc

    1228. 

  1228. 	nr.split=cur.vv-dev/n.split

    1229. 

  1229. #	print((nr.split*n.split+nr.desc*n.desc)/(n.desc+n.split)-cur.vv)

    1230. 

  1230. 	return(list(nr.desc=nr.desc, nr.split=nr.split))

    1231. 

  1231. }

    1232. 

  1232. 

    1233. 

  1233. split.value=function(cur.vv, n.desc, n.split, factor=log(2)){

    1234. 

  1234. 	dev=cur.vv-adjust.value(cur.vv, factor)

    1235. 

  1235. 	nr.desc=cur.vv + dev/n.desc

    1236. 

  1236. 	nr.split=cur.vv - dev/n.split

    1237. 

  1237. #	print((nr.split*n.split+nr.desc*n.desc)/(n.desc+n.split)-cur.vv)

    1238. 

  1238. 	return(list(nr.desc=nr.desc, nr.split=nr.split))

    1239. 

  1239. }

    1240. 

  1240. 

    1241. 

  1241. compare.rates=function(nodes=list(A=c(NULL,NULL),B=c(NULL,NULL)), phy, posterior.rates, ...){

    1242. 

  1242. 	if(is.null(names(nodes))) names(nodes)=c("A","B")

    1243. 

  1243. 	rates.a=rates[,match(nodes[[1]],names(rates))]

    1244. 

  1244. 	rates.b=rates[,match(nodes[[2]],names(rates))]

    1245. 

  1245. 	edges.a=phy$edge.length[match(nodes[[1]],phy$edge[,2])]

    1246. 

  1246. 	edges.b=phy$edge.length[match(nodes[[2]],phy$edge[,2])]

    1247. 

  1247. 	weights.a=edges.a/sum(edges.a)

    1248. 

  1248. 	weights.b=edges.b/sum(edges.b)

    1249. 

  1249. 	r.scl.a=apply(rates.a, 1,function(x) sum(x*weights.a))

    1250. 

  1250. 	r.scl.b=apply(rates.b, 1,function(x) sum(x*weights.b))

    1251. 

  1251. 	return(list(scl.A=r.scl.a, scl.B=r.scl.b, r.test=randomization.test(r.scl.a, r.scl.b, ...)))	

    1252. 

  1252. }

    1253. 

  1253. 

    1254. 

  1254. randomization.test=function(obs=obs, exp=exp, iter=10000, verbose=F, two.tailed=F){

    1255. 

  1255. 	if(verbose && length(obs)<iter) warning("iterations exceeds observed data")

    1256. 

  1256. 	if(verbose) warning(paste("p-value roughly approximated for ",length(exp), " values\n",sep=""))

    1257. 

  1257. 	

    1258. 

  1258. 	O=sample(obs, iter, replace=TRUE)

    1259. 

  1259. 	E=sample(exp, iter, replace=TRUE)

    1260. 

  1260. 	obj=data.frame(obs, exp)

    1261. 

  1261. 	names(obj)=c("obs","exp")

    1262. 

  1262. 	

    1263. 

  1263. 	result=list(c(O-E))

    1264. 

  1264. 	p=array(NA, iter)

    1265. 

  1265. 	

    1266. 

  1266. 	for(i in 1:length(result[[1]])){

    1267. 

  1267. 		if(result[[1]][i]>0) {

    1268. 

  1268. 			p[i]=1

    1269. 

  1269. 		} else {

    1270. 

  1270. 			if(result[[1]][i]==0) {

    1271. 

  1271. 				p[i]=sample(c(0,1),1)

    1272. 

  1272. 			} else {

    1273. 

  1273. 				p[i]=0

    1274. 

  1274. 			}

    1275. 

  1275. 		}

    1276. 

  1276. 	}

    1277. 

  1277. 	

    1278. 

  1278. 	p.g=round(1-(sum(p)/iter), digits=7)

    1279. 

  1279. 	p.l=1-p.g

    1280. 

  1280. 	

    1281. 

  1281. #	names(p.g)="greater"

    1282. 

  1282. #	names(p.l)="lesser"

    1283. 

  1283. 	

    1284. 

  1284. 	if(verbose) res=list(c(O-E),greater=p.g,lesser=p.l) else res=list(greater=p.g,lesser=p.l)

    1285. 

  1285. 	if(two.tailed) res=min(c(p.g,p.l))*2 else res=res

    1286. 

  1286. 	return(res)

    1287. 

  1287. }

    1288. 

  1288. 

    1289. 

  1289. collect.nodes=function(tips,phy,strict=FALSE){

    1290. 

  1290. 	if(length(tips)==1) stop("Must supply at least two tips.")

    1291. 

  1291. 	tt=phy$tip.label

    1292. 

  1292. 	tt=lapply(tips, function(x) which(phy$tip.label==x))

    1293. 

  1293. 	nodes=lapply(tt, function(x) get.ancestors.of.node(x, phy))

    1294. 

  1294. 	all.nodes=nodes[[1]]

    1295. 

  1295. 	for(i in 2:length(nodes)) {

    1296. 

  1296. 		all.nodes=intersect(all.nodes,nodes[[i]])

    1297. 

  1297. 	}

    1298. 

  1298. 	base.node=max(all.nodes)

    1299. 

  1299. 	if(strict) {

    1300. 

  1300. 		u=unlist(nodes)

    1301. 

  1301. 		return(c(sort(unique(u[u>=base.node])), unlist(tt)))

    1302. 

  1302. 	} else {

    1303. 

  1303. 		return(c(base.node,get.descendants.of.node(base.node,phy)))

    1304. 

  1304. 	}

    1305. 

  1305. }

    1306. 

  1306. 

    1307. 

  1307. find.relatives=function(tips,phy){

    1308. 

  1308. 	nn=collect.nodes(tips, phy, strict=FALSE)

    1309. 

  1309. 	return(phy$tip.label[nn[nn<=Ntip(phy)]])

    1310. 

  1310. }
    1311.