bd.shifts.optim(TreePar)
bd.shifts.optim()所属R语言包：TreePar
bd.shifts.optim: Estimating speciation and extinction rate changes and mass extinction events in phylogenies
bd.shifts.optim：，估计物种和物种灭绝速度的变化和在系统发育的生物大灭绝事件
译者：生物统计家园网 机器人LoveR
描述----------Description----------
bd.shifts.optim estimates the maximum likelihood speciation and extinction rates together with the rate shift times t=(t_1,t_2 .., t_m) in a (possibly incomplete sampled) phylogeny. At the times t, the rates are allowed to change and the species may undergo a mass extinction event.
bd.shifts.optim连同估计最大似然形态和消光率的速率移时间t =（T_1，T_2 ..，t_m），在一个（可能是不完整的采样）系统发育。在时间t的价格可以改变和物种可能进行大规模灭绝事件。
用法----------Usage----------
bd.shifts.optim(x, sampling, grid, start, end, maxitk = 5, yule = FALSE, ME = FALSE, all = FALSE, posdiv = FALSE, miniall = c(0), survival = 1,groups=0)
参数----------Arguments----------
参数：x
Vector of speciation times in the phylogeny. Time is measured increasing going into the past with the present being time 0. x can be obtained from a phylogenetic tree using getx(TREE).
向量的形态时代的亲缘关系。时间测量增加进入的过去与目前时间0。 x可以从系统树的getX（TREE）。
参数：sampling
Vector of length m. sampling_i is the probability of a species surviving the mass extinction at time t_i. sampling_1 is the probability of an extant species being sampled. sampling_1=1 means that the considered phylogeny is complete. sampling_i=1 (i>1) means that at time t_i, a rate shift may occur but no species go extinct. If ME=TRUE, all entries but sampling_1 will be discarded as they are estimated (however, input a vector sampling of the appropriate length such that the program knows how many mass extinction events you want to allow for).
向量的长度为m。 sampling_i时间t_i幸存的物种大灭绝一个物种的概率。 sampling_1是一个现存的物种被抽样的概率。 sampling_1 = 1表示所考虑的系统发育完成。 sampling_i = 1（I> 1）表示在时间t_i，一个变动可能发生，但没有物种灭绝。如果ME = TRUE，所有的作品，但sampling_1将被丢弃，因为他们估计（但是，输入适当的长度，这样的程序知道多少大规模灭绝事件，你要允许一个矢量采样）。
参数：grid, start, end
The model parameters are optimized for different fixed rate shift times. The fixed rate shift times are specified by being at (start, start+grid, start+2*grid .. end). I calculate the likelihood for the different rate shift times t instead of optimizing t with the function optim used for the other parameters, as the optimization performed poor for t (namely getting stuck in local optima).
的模型参数进行了优化不同的固定速率换档时间。固定利率转变的时间规定的是（启动，启动+网格，开始+2 *格..结束）。我计算不同的速率转变时间t的可能性，而不是用于其他参数的功能OPTIM优化吨，执行的优化差为T（即陷入局部最优）。
参数：yule
yule=TRUE sets the extinction rates to zero.
尤尔= TRUE设置灭绝利率降至零。
参数：maxitk
Integer value defining how many iterations shall be done in the optimization. Default is 5, but needs to be increased if too many warnings "convergence problem" appear.
整型值确定，应做多少次迭代的优化。默认值是5，但需要增加，如果出现太多警告“衔接的问题”。
参数：ME
ME=FALSE (default) uses the mass extinction fractions specified in sampling and does not estimate them. If ME=FALSE is used with sampling=c(1,1, .. , 1), no mass extinction events are considered.
ME = FALSE（默认）使用指定的采样的生物大灭绝分数并没有估计。如果ME = FALSE使用采样= C（1,1，...，1），被认为是没有大规模的生物灭绝事件。
参数：all
Only relevant when ME=TRUE. all=FALSE (default and recommended) estimates one speciation and one extinction rate for the whole tree, and estimates the intensities sampling_i (i>1) of mass extinction events. all=TRUE allows for varying speciation and extinction rates. Since the parameters might correlate, all=TRUE is not recommended.
只有有关时ME = TRUE。所有= FALSE（默认和推荐）估计，整个树形态和一种鸟类灭绝率，估计强度sampling_i（1）大灭绝事件。所有允许不同的形态和物种灭绝率= TRUE。由于相关的参数，所有= TRUE不推荐使用。
参数：posdiv
posdiv=FALSE (default) allows the speciation - extinction rate to be negative, i.e. allows for periods of declining diversity. posdiv=TRUE forces the speciation - extinction rate to be positive.
posdiv = FALSE（缺省）允许的形态 - 灭绝率是负的，即允许期间下降的多样性。 posdiv = TRUE强制的物种 - 物种灭绝速度是积极的。
参数：miniall
If you run the bd.shifts.optim for k shifts, but you now want to have K>k shifts, then set for continuing the analysis: update sampling, and set miniall=res[[2]] where res[[2]] is the output from the run with k shifts.
如果你运行bd.shifts.optim的对k的变化，但你现在想K> K表的变化，然后继续分析：更新采样，并设置miniall = [2]其中res [2] ]是具有k位移从运行的输出。
参数：survival
If survival = 1: The likelihood is conditioned on survival of the process (recommended). Otherwise survival = 0.
如果生存= 1：可能的条件是生存的过程（推荐）。否则生存= 0。
参数：groups
If groups != 0: the first column of groups indicates the age of higher taxa and the second column the number of species in the higher taxa (each row in groups corresponds to a leaf in the tree).
如果群体！= 0时，第一列的组，表示较高的分类群和第二列的物种数在较高的分类单元（组中的每一行对应于树中的一个叶）的年龄。
值----------Value----------
参数：res[[1]][[i]]
List of maximum likelihood parameter estimates for each fixed t where i-1 shifts are allowed to occur (i in 1:m).
最大似然参数估计列表，其中的i-1位移允许发生（i在1：米）的每个固定的t。
参数：res[[2]][[i]]
Maximum likelihood parameter estimates for i-1 shifts (i in 1:m): First entry is the (-log likelihood) value. The next i entries are the turnover (extinction/speciation) estimates, for the successive intervals going back in time. The next i entries are the diversification rate estimates (speciation-extinction). The next i-1 entries are the sampling estimates (if ME=TRUE). The last i-1 entries are the shift times. (Note: if ME=TRUE and all==FALSE, the second entry is the turnover, the third the diversification rate, followed by the sampling estimates).
最大似然参数估计的i-1的变化（1米）：第一项是（对数似然）值。接下来，我项目的营业额（消光/形态）的估计，连续间隔的时间。接下来，我的项目是多元化率估计值（物种灭绝）。 I-1项目的抽样估计（ME = TRUE）。最后的I-1项目是换档时间。 （注：如果ME = == TRUE和FALSE，第二个项目的营业额，第三个多元化率，采样估计）。
参数：res[[3]]
Vector of time points where the function was evaluated.
其中的功能进行评价的时间点的向量。
参数：res[[4]]
Array specifying the time points when there was a convergence problem: a row of res[[4]] with entry (i,t_i) means that when adding the i-th shift at time t_i, a convergence problem was encountered.
阵列指定的时间点，当有一个收敛的问题：一排的res [[4]]与的条目（ⅰt_i）指，当加入的第i个移位时t_i，收敛时遇到问题。
注意----------Note----------
The likelihood is calculated assuming there were two lineages at the time of the root. The likelihood is conditioned on survival of the two lineages if survival = 1. Likelihood-values from bd.densdep.optim are directly comparable (eg AIC) for survival = 0. Likelihood-values from laser are comparable for survival = 0 after the transformation -res+(sum(log(2:length(x)))-(length(x)-1)*log(2)).
的可能性来计算假设有两个谱系，在根的时间。的可能性的条件是生存的生存两个谱系，如果= 1。似然值来自bd.densdep.optim的直接可比性（如AIC）= 0的生存。似然值从激光相媲美的生存= 0，改造后的水库+（SUM（log（2：长度（x））） - （长度（x）-1）* LOG（2））。
（作者）----------Author(s)----------
Tanja Stadler
参考文献----------References----------
T. Stadler & F. Bokma: Estimating speciation and extinction rates for phylogenies of higher taxa. Submitted, 2012. (for groups>0 but no rate shift) A. Lambert & T. Stadler: Coalescent point processes and phylogenies. Manuscript. (for groups>0)
实例----------Examples----------
set.seed(1)
# First we simulate a tree, and then estimate the parameters for the tree:[首先，我们模拟了树，然后估计的参数树：]
# Number of species[种数]
nspecies <- 20
# At time 1 and 2 in the past, we have a rate shift:[在过去，在时刻1和2中，我们有一个变动：]
time <- c(0,1,2)
# Mass extinction intensities 0.5 at time 1 in past, 0.4 at time 2 in past. Present day species are all sampled (rho_1=1):[大灭绝的强度在过去的时间过去，0.4时间1 0.5。现今的物种所有样本（rho_1，= 1）：]
rho <- c(1,0.5,0.4)
# speciation rates (between t_i,t_{i+1} we have speciation rate lambda_i):[形态率（在t_i，T_ {i +1}我们有形态率lambda_i），]
lambda <- c(2,2,1)
# extinction rates (between t_i,t_{i+1} we have extinction rate mu_i):[灭绝率（在t_i，T_ {i +1}，我们有灭绝的速度mu_i）：]
mu <- c(1,1,0)
# Simulation of a tree:[模拟一棵树：]
tree<-sim.rateshift.taxa(nspecies,1,lambda,mu,frac=rho,times=time,complete=FALSE)
# Extracting the speciation times x:[提取的形态时间：]
x<-sort(getx(tree[[1]][[1]]),decreasing=TRUE)
# When estimating the shift times t for x, we allow the shift times to be 0.6, 0.8, 1, 1.2, .. ,2.4:[当推定变速时间t为x时，我们允许换档时间是0.6，0.8，1 1.2，.. 2.4：]
start <- 0.6
end <- 2.4
grid <- 0.2
# We fix rho and estimate time, lambda, mu:[我们修复，LAMBDA，亩rho和估计时间：]
res <- bd.shifts.optim(x,rho,grid,start,end)
res[[2]]
# We fix rho=1 and mu=0 and then estimate time, lambda:[我们解决ρ= 1亩= 0，然后估计时间，λ：]
resyule <- bd.shifts.optim(x,rho,grid,start,end,yule=TRUE)
resyule[[2]]
# We estimate time, rho, lambda, mu:[我们估计，ρ，λ，亩：]
resrho <- bd.shifts.optim(x,rho,grid,start,end,ME=TRUE)
resrho[[2]]
# Data analysis in Stadler & Bokma, 2012:[数据分析施泰德和Bokma，2012年：]
# Number of species in each order from Sibley and Monroe (1990)[在每个订单的种数从西布利和梦露（1990）]
data(bird.orders)
S <- c(10, 47, 69, 214, 161, 17, 355, 51, 56, 10, 39, 152,6, 143, 358, 103, 319, 23, 291, 313, 196, 1027, 5712)
groups<-get.groups(bird.orders,S,0)
groupscut<-get.groups(bird.orders,S,96.43*0.207407)
x<-branching.times(bird.orders)
bd.shifts.optim(x,sampling=c(1),survival=1,groups=groups)[[2]]
bd.shifts.optim(x,sampling=c(1),survival=1,groups=groupscut)[[2]]
# allowing one shift in rates:[允许转移率：]
bd.shifts.optim(x,sampling=c(1,1),grid=1,start=20,end=25,survival=1,groups=groupscut)[[2]]
转载请注明:出自 生物统计家园网(http://www.biostatistic.net)。
注：
注1：为了方便大家学习，本文档为生物统计家园网机器人LoveR翻译而成，仅供个人R语言学习参考使用，生物统计家园保留版权。
注2：由于是机器人自动翻译，难免有不准确之处，使用时仔细对照中、英文内容进行反复理解，可以帮助R语言的学习。
注3：如遇到不准确之处，请在本贴的后面进行回帖，我们会逐渐进行修订。
|
发表于 2012-10-1 11:51:10
