2

我正在使用包中的flexmixedruns函数，当我尝试运行该函数fpc时得到一个 。flexmix error

我的数据集既有连续数据又有分类数据，但我的大多数分类数据只有 2 个级别（"Y"或"N"）。我的一些变量有几个级别。我想知道我是否收到错误，因为该函数将我所有的分类变量都视为多项分布。

有没有人有使用此功能的经验？

可重现的例子

#####  check for fpc package
required.packages <- c("fpc")
new.packages <- 
  required.packages[!(required.packages %in% installed.packages()[,"Package"])] 
if(length(new.packages)) install.packages(new.packages)
rm(required.packages, new.packages)

library(fpc)

#####  create data set
df <- matrix(
  data=c("widget1", "widget2", "widget3", "widget4", "widget5", "widget6",
                    58, 18, 31, 130, 40, 31, 
                    70, 19, 44, 120, 57, 50,
                    "1E6", "1E5", "1E4", "1E6", "1E5", "1E4",
                    "Y", "Y", "N", "N", "N", "Y",
                    "N", "Y", "N", "Y", "N", "Y"), 
             nrow=6, ncol=6)

df <- as.data.frame(x=df)
row.names(df) <- df[, 1]
df <- df[, -1]
colnames(df) <- c("cont1", "cont2", "multi1", "bin1", "bin2")

df$cont1 <- as.numeric(df$cont1)
df$cont2 <- as.numeric(df$cont2)

#####  model
mdl <-
  flexmixedruns(x=df, xvarsorted=TRUE, continuous=2, discrete=3, simruns=5,
                n.cluster=3, recode=TRUE)

错误信息

Error in summary(flexout[[optimalk]]) : 
    error in evaluating the argument 'object' in selecting a method for function 'summary': Error in flexout[[optimalk]] : attempt to select less than one element
4

2 回答 2

0

如果没有更详细的调试，就无法确定。确保加载了方法包，即

library(methods)
于 2014-09-10T01:46:54.533 回答
0

问题是数据格式。该函数最初似乎无法处理因子数据，尽管文档说它可以对可以转换为因子变量的分类数据进行任何处理。

df$multi1 = as.numeric(df$multi1)
df$bin1 = as.numeric(df$bin1)
df$bin2 = as.numeric(df$bin2)

mdl <-flexmixedruns(x=df, xvarsorted=TRUE, continuous=2, discrete=3, simruns=5,
                n.cluster=3, recode=TRUE)

summary(df)
     cont1       cont2         multi1       bin1          bin2    
 Min.   :1   Min.   :1.0   Min.   :1   Min.   :1.0   Min.   :1.0  
 1st Qu.:2   1st Qu.:2.0   1st Qu.:1   1st Qu.:1.0   1st Qu.:1.0  
 Median :3   Median :3.5   Median :2   Median :1.5   Median :1.5  
 Mean   :3   Mean   :3.5   Mean   :2   Mean   :1.5   Mean   :1.5  
 3rd Qu.:4   3rd Qu.:5.0   3rd Qu.:3   3rd Qu.:2.0   3rd Qu.:2.0  
 Max.   :5   Max.   :6.0   Max.   :3   Max.   :2.0   Max.   :2.0 

# k=  3  new best fit found in run  1 
# Nonoptimal or repeated fit found in run  2 
# k=  3  new best fit found in run  3 
# Nonoptimal or repeated fit found in run  4 
# Nonoptimal or repeated fit found in run  5 
# k=  3  BIC=  216462.2
于 2017-03-07T21:44:19.373 回答