The dataframe is visible with out any error. But when the same is printed using write.xlsx fucnction of the package XLSX, it gives the error.
可以看到dataframe有任何错误。但是当同样的东西用写来打印时。xlsx软件包的函数,它给出了错误。
Error in .jcall(cell, "V", "setCellValue", value) :
method setCellValue with signature ([D)V not found.
The dput of the data.frame looks like:
数据的dput。框架如下:
Timestamp qs pqs logqs es p_imp dep r_dep agg_rtn
(time) (dbl) (dbl) (dbl) (dbl) (dbl) (dbl) (dbl) (dbl)
1 2015-05-04 09:29:59 0.05788732 0.0007478696 0.0007478545 0.09633803 -0.0446830986 3533.518 274079.9 -0.0006432937
2 2015-05-04 10:00:00 0.04948394 0.0006362707 0.0006362707 0.07586009 0.0088016055 2416.431 187953.1 0.0000000000
3 2015-05-04 10:30:00 0.05554795 0.0007142532 0.0007142532 0.06417808 -0.0002739726 3245.574 252422.0 0.0000000000
4 2015-05-04 10:59:59 0.04863014 0.0006194244 0.0006194244 0.08434442 0.0024951076 3563.401 279503.9 0.0000000000
5 2015-05-04 11:30:00 0.05761986 0.0007319037 0.0007319037 0.07851027 0.0154965753 2010.943 158429.1 -0.0006339144
6 2015-05-04 12:00:00 0.04957627 0.0006285051 0.0006285051 0.07025424 0.0070762712 1819.908 143546.0 0.0000000000
Variables not shown: vol_30_sum (dbl), vol_30_mean (dbl), p_return_sqr (dbl), p_return_mean (dbl), Lim_or_out (dbl),
closing_price (dbl), closing_vol (dbl)
Kindly help in resolving this error.
请帮忙解决这个错误。
3 个解决方案
#1
37
Still no reproducible example, but from your class(q1)
it appears that q1
is a tbl_df
(the sort of dataframe that the dplyr
package produces) whereas write.xlsx
expects a data.frame
.
仍然没有可复制的示例,但是从您的类(q1)中可以看出q1是tbl_df (dplyr包产生的那种数据aframe),而write则是q1。xlsx预计data.frame。
Try giving write.xlsx
a plain data.frame
as it expects. e.g.
试着写。xlsx是一个普通的数据。如。
write.xlsx(as.data.frame(q1), ...)
Here's a reproducible example (i.e. you could copy-paste it into your R session to reproduce the bug + fix).
这里有一个可复制的示例(例如,您可以将它复制到R会话中,以复制bug + fix)。
library(dplyr)
iris2 <- tbl_df(iris)
class(iris2) # like yours
# [1] "tbl_df" "tbl" "data.frame"
# Now let's try to write to XLSX using command as mentioned in your comments
library(xlsx)
write.xlsx(iris2, file='test.xlsx', sheetName="Sheet1", col.names=TRUE, row.names=FALSE, append=TRUE)
# Error in .jcall(cell, "V", "setCellValue", value) :
# method setCellValue with signature ([D)V not found
# In addition: Warning message:
# In if (is.na(value)) { :
# the condition has length > 1 and only the first element will be used
# ^--- we can reproduce your error. This is the point of a reproducible example, so we can see if our fixes work for you.
Now let's try fix it by making sure that write.xlsx gets a data.frame, not a tbl_df!
现在我们试着把它写下来。xlsx获得的是数据。帧,而不是tbl_df!
write.xlsx(as.data.frame(iris2), file='test.xlsx', sheetName="Sheet1", col.names=TRUE, row.names=FALSE, append=TRUE)
# huzzah!
#2
3
I find this happens when grouping variables with dplyr. If you end a chain with %>% ungroup () it appears to resolve
当用dplyr对变量进行分组时,我发现会发生这种情况。如果您用%>% ungroup()结束一个链,那么它似乎可以解析
#3
0
It seems that there is a bug with the Date/Time format of the first column (Timestamp). If you convert the first column to character, it should work. So, you may change your first column to
似乎第一列(时间戳)的日期/时间格式有错误。如果您将第一列转换为字符,它应该可以工作。所以,你可以改变你的第一列。
q1[,1] <- as.character(q1[,1])
and try again...
再试一次…
#1
37
Still no reproducible example, but from your class(q1)
it appears that q1
is a tbl_df
(the sort of dataframe that the dplyr
package produces) whereas write.xlsx
expects a data.frame
.
仍然没有可复制的示例,但是从您的类(q1)中可以看出q1是tbl_df (dplyr包产生的那种数据aframe),而write则是q1。xlsx预计data.frame。
Try giving write.xlsx
a plain data.frame
as it expects. e.g.
试着写。xlsx是一个普通的数据。如。
write.xlsx(as.data.frame(q1), ...)
Here's a reproducible example (i.e. you could copy-paste it into your R session to reproduce the bug + fix).
这里有一个可复制的示例(例如,您可以将它复制到R会话中,以复制bug + fix)。
library(dplyr)
iris2 <- tbl_df(iris)
class(iris2) # like yours
# [1] "tbl_df" "tbl" "data.frame"
# Now let's try to write to XLSX using command as mentioned in your comments
library(xlsx)
write.xlsx(iris2, file='test.xlsx', sheetName="Sheet1", col.names=TRUE, row.names=FALSE, append=TRUE)
# Error in .jcall(cell, "V", "setCellValue", value) :
# method setCellValue with signature ([D)V not found
# In addition: Warning message:
# In if (is.na(value)) { :
# the condition has length > 1 and only the first element will be used
# ^--- we can reproduce your error. This is the point of a reproducible example, so we can see if our fixes work for you.
Now let's try fix it by making sure that write.xlsx gets a data.frame, not a tbl_df!
现在我们试着把它写下来。xlsx获得的是数据。帧,而不是tbl_df!
write.xlsx(as.data.frame(iris2), file='test.xlsx', sheetName="Sheet1", col.names=TRUE, row.names=FALSE, append=TRUE)
# huzzah!
#2
3
I find this happens when grouping variables with dplyr. If you end a chain with %>% ungroup () it appears to resolve
当用dplyr对变量进行分组时,我发现会发生这种情况。如果您用%>% ungroup()结束一个链,那么它似乎可以解析
#3
0
It seems that there is a bug with the Date/Time format of the first column (Timestamp). If you convert the first column to character, it should work. So, you may change your first column to
似乎第一列(时间戳)的日期/时间格式有错误。如果您将第一列转换为字符,它应该可以工作。所以,你可以改变你的第一列。
q1[,1] <- as.character(q1[,1])
and try again...
再试一次…