我试图使用Bray-Curtis相异度使用来自包 vegan 的metaMDS进行NMDS测试。请参见代码：
pc = read.csv("nmds.csv", header = TRUE) com = pc[,5:ncol(pc)] ###其中5是我的信息开始的列m_com = as.matrix(com) set.seed(123) nmds = metaMDS(m_com, distance = "bray")
这就是我的数据看起来的样子：data
然而，当我执行最后一步（nmds = metaMDS（m_com，distance =“bray”））时，我得到以下错误：Error in if (max(dis) > maxdis + sqrt(.Machine$double.eps)) { : missing value where TRUE/FALSE needed In addition: Warning messages: 1: In distfun(comm, method = distance, ...) : you have empty rows: their dissimilarities may be meaningless in method “bray” 2: In distfun(comm, method = distance, ...) : missing values in results
我已经使用相同的代码多年，从来没有遇到过这个问题。我在一些线程中看到，这是因为我的数据中有零，但我以前对包含零的数据做过这样的处理。
我在一组包含许多零的旧数据上使用了相同的代码（但数据集要小得多），它能够计算它并给予解决方案（见下面的结果）。

Wisconsin double standardization Run 0 stress 0.1232691  Run 1 stress
0.1380158  Run 2 stress 0.1232695  ... Procrustes: rmse 0.0002867434  max resid 0.002609401  ... Similar to previous best Run 3 stress
0.1232695  ... Procrustes: rmse 0.0002975044  max resid 0.002720978  ... Similar to previous best Run 4 stress 0.2390168  Run 5 stress
0.1232696  ... Procrustes: rmse 0.0002939605  max resid 0.002594778  ... Similar to previous best Run 6 stress 0.1502755  Run 7 stress
0.1232693  ... Procrustes: rmse 0.0002335296  max resid 0.002429394  ... Similar to previous best Run 8 stress 0.1232691  ... New best solution ... Procrustes: rmse 2.084762e-06  max resid 1.672081e-05  ... Similar to previous best Run 9 stress 0.1232695  ... Procrustes: rmse 0.0002904649  max resid 0.002643373  ... Similar to previous best Run 10 stress 0.1380158  Run 11 stress 0.1232695  ... Procrustes: rmse
0.0002872963  max resid 0.002608208  ... Similar to previous best Run 12 stress 0.1232695  ... Procrustes: rmse 0.0002861628  max resid
0.002597991  ... Similar to previous best Run 13 stress 0.1232696  ... Procrustes: rmse 0.000292253  max resid 0.002573828  ... Similar to previous best Run 14 stress 0.2587088  Run 15 stress 0.1232695  ... Procrustes: rmse 0.0002856807  max resid 0.002589665  ... Similar to previous best Run 16 stress 0.1232695  ... Procrustes: rmse
0.0002866945  max resid 0.002595448  ... Similar to previous best Run 17 stress 0.1232695  ... Procrustes: rmse 0.0002864601  max resid
0.002605703  ... Similar to previous best Run 18 stress 0.1232695  ... Procrustes: rmse 0.000295793  max resid 0.002704928  ... Similar to previous best Run 19 stress 0.1232695  ... Procrustes: rmse
0.0002905679  max resid 0.002632124  ... Similar to previous best Run 20 stress 0.1232695  ... Procrustes: rmse 0.0002865427  max resid
0.002612774  ... Similar to previous best
*** Best solution repeated 11 times
> nmds

Call: metaMDS(comm = m_com, distance = "bray") 

global Multidimensional Scaling using monoMDS

Data:     wisconsin(m_com)  Distance: bray 

Dimensions: 2  Stress:     0.1232691  Stress type 1, weak ties 
Best solution was repeated 11 times in 20 tries 
The best solution was from try 8 (random start) 
Scaling: centring, PC rotation, halfchange scaling  
Species: expanded scores based on ‘wisconsin(m_com)’

为什么它不能在我的新数据集上工作？我已经验证了所有框都包含一个值，即使它是零。我很困惑，为什么它只在一个数据集上工作，而不是另一个。会不会是因为文件的大小？任何帮助都很感激。谢谢大家。