logo标签列表

R中从长格式转换为宽格式时出现问题

rdataframe
4

在查看不同的线程时，我了解到 stats 包中的函数 reshape。我在一个“虚拟”数据集上使用它没有问题，并成功将其从长格式转换为宽格式。然而，我不知道为什么它在我的数据上不起作用，因为它和我使用的对象几乎相同，数据类型也相似。我希望你们能帮助我找出它行为异常的原因。

无论如何，这个没有问题：

> df <- data.frame(
  +     year   = c(rep(2000, 12), rep(2001, 12)),
  +     month  = rep(1:12, 2),
  +     values = rnorm(24)
  + )

# year    month    values
1  2000     1    1.52435428
2  2000     2   -0.89394797
3  2000     3    0.75965499
4  2000     4    1.21497443

转换为宽格式：


df_wide <- reshape(df, idvar="year", timevar="month", v.names="values", direction="wide")

# year    values_1  values_2  values_3  values_4   values_5    values_6   values_7 values_8  values_9  values_10 values_11  values_12
1  2000 1.524354 -0.893948  0.759655 1.2149744 -1.3237634 -0.08681768  0.5208436 -0.2602807 0.6378904 -0.9852600 -1.128048 -0.1466028
2  2001 1.913969 -1.966720 -0.947688 0.8375891 -0.1015944  1.11812723 -1.5164472 -0.7089485 0.5975851  0.2514546 -1.578210 -0.9044418

但是当使用我的数据时，它看起来就像这样：

my_df <- dput(head(experiment, 30))

structure(list(transcript = c("TR100743-c0_g1_i3", "TR100743-c0_g1_i3", 
"TR100743-c0_g1_i3", "TR100743-c0_g1_i3", "TR100743-c0_g1_i3", 
"TR100987-c0_g1_i2", "TR100987-c0_g1_i2", "TR100987-c0_g1_i2", 
"TR100987-c0_g1_i2", "TR100987-c0_g1_i2", "TR101301-c4_g1_i16", 
"TR101301-c4_g1_i16", "TR101301-c4_g1_i16", "TR101301-c4_g1_i16", 
"TR101301-c4_g1_i16", "TR102190-c1_g1_i1", "TR102190-c1_g1_i1", 
"TR102190-c1_g1_i1", "TR102190-c1_g1_i1", "TR102190-c1_g1_i1", 
"TR102346-c0_g2_i1", "TR102346-c0_g2_i1", "TR102346-c0_g2_i1", 
"TR102346-c0_g2_i1", "TR102346-c0_g2_i1", "TR102352-c4_g2_i5", 
"TR102352-c4_g2_i5", "TR102352-c4_g2_i5", "TR102352-c4_g2_i5", 
"TR102352-c4_g2_i5"), hours = c(0, 2, 8, 24, 48, 0, 2, 8, 24, 
48, 0, 2, 8, 24, 48, 0, 2, 8, 24, 48, 0, 2, 8, 24, 48, 0, 2, 
8, 24, 48), exp.change = c(NA, -43.1958273184645, -61.3014008509066, 
964.925115099619, -52.7060728326392, NA, -46.2563848585369, 3.29396898799807, 
-99.9994681489801, 106710484.025972, NA, -29.6341333478577, 522.224859380388, 
40.4737694947169, -1.34388206141046, NA, -18.7670826937756, 5.49472822880452, 
55.1072690537026, 33.5824607349752, NA, -99.999962131178, 789697313.24393, 
18.6337471833012, 52.4442959208125, NA, -31.3334122297108, 9.64745757892995, 
28.48552519881, 70.5808772231999), response = c("Primary", "Primary", 
"Primary", "Primary", "Primary", "Primary", "Primary", "Primary", 
"Primary", "Primary", "Primary", "Primary", "Primary", "Primary", 
"Primary", "Tertiary", "Tertiary", "Tertiary", "Tertiary", "Tertiary", 
"Primary", "Primary", "Primary", "Primary", "Primary", "Primary", 
"Primary", "Primary", "Primary", "Primary")), class = c("tbl_df", 
"tbl", "data.frame"), row.names = c(NA, -30L))


# transcript        hours   exp.change response   
 1 TR100743-c0_g1_i3     0        NA    Primary 
 2 TR100743-c0_g1_i3     2       -43.2  Primary 
 3 TR100743-c0_g1_i3     8       -61.3  Primary 
 4 TR100743-c0_g1_i3    24       965.   Primary 
 5 TR100743-c0_g1_i3    48       -52.7  Primary 
 6 TR100987-c0_g1_i2     0        NA    Primary 
 7 TR100987-c0_g1_i2     2       -46.3  Primary 
 8 TR100987-c0_g1_i2     8         3.29 Primary 
 9 TR100987-c0_g1_i2    24      -100.0  Primary 
10 TR100987-c0_g1_i2    48 106710484.   Primary

当我尝试“重塑”它时，出现了这个错误：

my_df_wide <- reshape(my_df, idvar = c("transcript", "response"), timevar = "hours", v.names="exp.change", direction = "wide")

# transcript         response  `exp.change.c(0, 2, 8, 24, 48)`
 1 TR100743-c0_g1_i3  Primary                                NA
 2 TR100987-c0_g1_i2  Primary                                NA
 3 TR101301-c4_g1_i16 Primary                                NA
 4 TR102190-c1_g1_i1  Tertiary                               NA
 5 TR102346-c0_g2_i1  Primary                                NA
 6 TR102352-c4_g2_i5  Primary                                NA
 7 TR10396-c0_g1_i6   Primary                                NA
 8 TR11844-c0_g2_i1   Secondary                              NA
 9 TR12672-c1_g2_i1   Primary                                NA
10 TR12672-c1_g2_i2   Primary                                NA

是因为空值吗？我真的不知道为什么它会表现得这样...非常感谢任何帮助。

1
您希望调整后的输出是什么样子？ 类似这样吗？ `library(tidyverse);my_df %>% spread(hours, exp.change, sep = "_")` - AntoniosK
2个回答
6
使用stats::reshape重塑数据可能会很繁琐。Hadley Wickham及其团队花费了相当长的时间创建了一个全面的解决方案。首先是reshape2包，然后是tidyr中的spread()和gather()，现在这些已经被pivot_wider()和pivot_longer()所取代。以下是您可以使用tidyr::pivot_wider()实现所需结果的方法。
library(tidyr)
pivot_wider(
  my_df,
  id_cols = c(transcript, response),
  names_from = hours,
  values_from = exp.change,
  names_prefix = "exp.change_"
)
#> # A tibble: 6 x 7
#>   transcript response exp.change_0 exp.change_2 exp.change_8 exp.change_24
#>   <chr>      <chr>           <dbl>        <dbl>        <dbl>         <dbl>
#> 1 TR100743-… Primary            NA        -43.2       -61.3          965. 
#> 2 TR100987-… Primary            NA        -46.3         3.29        -100. 
#> 3 TR101301-… Primary            NA        -29.6       522.            40.5
#> 4 TR102190-… Tertiary           NA        -18.8         5.49          55.1
#> 5 TR102346-… Primary            NA       -100.  789697313.            18.6
#> 6 TR102352-… Primary            NA        -31.3         9.65          28.5
#> # … with 1 more variable: exp.change_48 <dbl>

我认为对于这两种转换（宽/长），tidyr命令有专门的命令和文档，相比之下，与stats :: reshape()相比，tidyr命令更容易使用。

编辑： stats::reshape()产生了奇怪的结果，因为似乎处理tibble格式的数据框时出现了问题。除此之外，您的命令是没问题的。只需添加一个as.data.frame()函数就可以解决问题。

reshape(
  as.data.frame(my_df),
  idvar = c("transcript", "response"),
  timevar   = "hours",
  v.names = "exp.change",
  direction = "wide"
)
#>            transcript response exp.change.0 exp.change.2  exp.change.8
#> 1   TR100743-c0_g1_i3  Primary           NA    -43.19583 -6.130140e+01
#> 6   TR100987-c0_g1_i2  Primary           NA    -46.25638  3.293969e+00
#> 11 TR101301-c4_g1_i16  Primary           NA    -29.63413  5.222249e+02
#> 16  TR102190-c1_g1_i1 Tertiary           NA    -18.76708  5.494728e+00
#> 21  TR102346-c0_g2_i1  Primary           NA    -99.99996  7.896973e+08
#> 26  TR102352-c4_g2_i5  Primary           NA    -31.33341  9.647458e+00
#>    exp.change.24 exp.change.48
#> 1      964.92512 -5.270607e+01
#> 6      -99.99947  1.067105e+08
#> 11      40.47377 -1.343882e+00
#> 16      55.10727  3.358246e+01
#> 21      18.63375  5.244430e+01
#> 26      28.48553  7.058088e+01

但由于您似乎已经在使用tidyverse，tidyr::pivot_wider()似乎是最适合的选择。

1
你也可以在 pivot_wider 中添加 names_sep = "_"，以获得更具信息性的列名，类似于 reshape 方法。 - AntoniosK
1
好的观点！然而，仅使用 names_sep = "_" 并不能解决这个问题。似乎 names_prefix = "exp.change_" 才是你所建议的。感谢你提出这个问题，我现在会进行编辑。 - Till
是的，看起来我必须将我的tibble转换为dataframe，然后就可以了！这样做让它工作了...是的，从现在开始我会尽量只使用tidyverse函数。 - John Sandman
2
使用stats::reshape 
# `idvar` has 1 value here 
reshape(my_df, idvar="transcript", timevar="hours", v.names="exp.change", direction="wide")

      transcript response exp.change.0 exp.change.2  exp.change.8 exp.change.24 exp.change.48
 1   TR100743-c0_g1_i3  Primary           NA    -43.19583 -6.130140e+01     964.92512 -5.270607e+01
 6   TR100987-c0_g1_i2  Primary           NA    -46.25638  3.293969e+00     -99.99947  1.067105e+08
11 TR101301-c4_g1_i16  Primary           NA    -29.63413  5.222249e+02      40.47377 -1.343882e+00
16  TR102190-c1_g1_i1 Tertiary           NA    -18.76708  5.494728e+00      55.10727  3.358246e+01
21  TR102346-c0_g2_i1  Primary           NA    -99.99996  7.896973e+08      18.63375  5.244430e+01
26  TR102352-c4_g2_i5  Primary           NA    -31.33341  9.647458e+00      28.48553  7.058088e+01

使用 data.table

setDT(my_df)
dcast(my_df,transcript~hours, value.var="exp.change")

           transcript  0         2             8        24            48
 1:  TR100743-c0_g1_i3 NA -43.19583 -6.130140e+01 964.92512 -5.270607e+01
 2:  TR100987-c0_g1_i2 NA -46.25638  3.293969e+00 -99.99947  1.067105e+08
 3: TR101301-c4_g1_i16 NA -29.63413  5.222249e+02  40.47377 -1.343882e+00
 4:  TR102190-c1_g1_i1 NA -18.76708  5.494728e+00  55.10727  3.358246e+01
 5:  TR102346-c0_g2_i1 NA -99.99996  7.896973e+08  18.63375  5.244430e+01
 6:  TR102352-c4_g2_i5 NA -31.33341  9.647458e+00  28.48553  7.058088e+01

 dcast(my_df,transcript + response ~hours, value.var="exp.change")

           transcript response  0         2             8        24            48
1:  TR100743-c0_g1_i3  Primary NA -43.19583 -6.130140e+01 964.92512 -5.270607e+01
2:  TR100987-c0_g1_i2  Primary NA -46.25638  3.293969e+00 -99.99947  1.067105e+08
3: TR101301-c4_g1_i16  Primary NA -29.63413  5.222249e+02  40.47377 -1.343882e+00
4:  TR102190-c1_g1_i1 Tertiary NA -18.76708  5.494728e+00  55.10727  3.358246e+01
5:  TR102346-c0_g2_i1  Primary NA -99.99996  7.896973e+08  18.63375  5.244430e+01
6:  TR102352-c4_g2_i5  Primary NA -31.33341  9.647458e+00  28.48553  7.058088e+01

您也可以使用旧的reshape2包。
 reshape2::dcast(my_df,transcript + response ~hours, value.var="exp.change")

          transcript response  0         2             8        24            48
1  TR100743-c0_g1_i3  Primary NA -43.19583 -6.130140e+01 964.92512 -5.270607e+01
2  TR100987-c0_g1_i2  Primary NA -46.25638  3.293969e+00 -99.99947  1.067105e+08
3 TR101301-c4_g1_i16  Primary NA -29.63413  5.222249e+02  40.47377 -1.343882e+00
4  TR102190-c1_g1_i1 Tertiary NA -18.76708  5.494728e+00  55.10727  3.358246e+01
5  TR102346-c0_g2_i1  Primary NA -99.99996  7.896973e+08  18.63375  5.244430e+01
6  TR102352-c4_g2_i5  Primary NA -31.33341  9.647458e+00  28.48553  7.058088e+01
网页内容由stack overflow 提供, 点击上面的
可以查看英文原文，
原文链接