當讀取下面的文本時,fread()
未能檢測到列8和9中的缺失值。這僅適用於默認選項integer64="integer64"
。設置integer64="double"
或"character"
正確檢測到NA
s。請注意,該文件在V8和V9中有三種可能的NAs-- ,,
; , ,
;和NA
。追加na.strings=c("NA","N/A",""," "), sep=","
作爲選項不起作用。fread()失敗,整數64列中的值缺失
使用read.csv()
的作用方式與fread(integer="double")
的作用相同。要被讀取
文本(也available as a file integer64_and_NA.csv):
2012,276,,0,"S1","001",1,,724135215,1590915056,
2012,276,2,8,"S1","001",1, ,,154598,0
2012,276,2,12,"S1","001",1,NA,5118863,21819477,
2012,276,2,0,"S1","011",8,3127133583,3127133583,9003982501,0
下面是從fread()
輸出:
DT <- fread(input="integer64_and_NA.csv", verbose=TRUE, integer64="integer64", na.strings=c("NA","N/A",""," "), sep=",")
Input contains no \n. Taking this to be a filename to open
Detected eol as \r\n (CRLF) in that order, the Windows standard.
Looking for supplied sep ',' on line 4 (the last non blank line in the first 'autostart') ... found ok
Found 11 columns
First row with 11 fields occurs on line 1 (either column names or first row of data)
Some fields on line 1 are not type character (or are empty). Treating as a data row and using default column names.
Count of eol after first data row: 5
Subtracted 1 for last eol and any trailing empty lines, leaving 4 data rows
Type codes: 11114412221 (first 5 rows)
Type codes: 11114412221 (after applying colClasses and integer64)
Type codes: 11114412221 (after applying drop or select (if supplied)
Allocating 11 column slots (11 - 0 NULL)
0.000s ( 0%) Memory map (rerun may be quicker)
0.000s ( 0%) sep and header detection
0.000s ( 0%) Count rows (wc -l)
0.000s ( 0%) Column type detection (first, middle and last 5 rows)
0.000s ( 0%) Allocation of 4x11 result (xMB) in RAM
0.000s ( 0%) Reading data
0.000s ( 0%) Allocation for type bumps (if any), including gc time if triggered
0.000s ( 0%) Coercing data already read in type bumps (if any)
0.000s ( 0%) Changing na.strings to NA
0.001s Total
所得data.table是:
DT
V1 V2 V3 V4 V5 V6 V7 V8 V9 V10 V11
1: 2012 276 NA 0 S1 001 1 9218868437227407266 724135215 1590915056 NA
2: 2012 276 2 8 S1 001 1 9218868437227407266 9218868437227407266 154598 0
3: 2012 276 2 12 S1 001 1 9218868437227407266 5118863 21819477 NA
4: 2012 276 2 0 S1 011 8 3127133583 3127133583 9003982501 0
NA
值是在不是的列中正確檢測到。對於V8和V9,其中fread()
標記爲整數64,而不是NAs,我們有「9218868437227407266」。 有趣的是,str()
返回V8和V9的相應值NA
:
str(DT)
Classes ‘data.table’ and 'data.frame': 4 obs. of 11 variables:
$ V1 : int 2012 2012 2012 2012
$ V2 : int 276 276 276 276
$ V3 : int NA 2 2 2
$ V4 : int 0 8 12 0
$ V5 : chr "S1" "S1" "S1" "S1"
$ V6 : chr "001" "001" "001" "011"
$ V7 : int 1 1 1 8
$ V8 :Class 'integer64' num [1:4] NA NA NA 1.55e-314
$ V9 :Class 'integer64' num [1:4] 3.58e-315 NA 2.53e-317 1.55e-314
$ V10:Class 'integer64' num [1:4] 7.86e-315 7.64e-319 1.08e-316 4.45e-314
$ V11: int NA 0 NA 0
- attr(*, ".internal.selfref")=<externalptr>
...但沒有別的把它們看作NA
:
is.na(DT$V8)
[1] FALSE FALSE FALSE FALSE
max(DT$V8)
integer64
[1] 9218868437227407266
> max(DT$V8, na.rm=TRUE)
integer64
[1] 9218868437227407266
> class(DT$V8)
[1] "integer64"
> typeof(DT$V8)
[1] "double"
它似乎並沒有成爲一個打印/只有屏幕問題,data.table
將它們視爲巨大整數:
DT[, V12:=as.numeric(V8)]
Warning message:
In as.double.integer64(V8) :
integer precision lost while converting to double
> DT
V1 V2 V3 V4 V5 V6 V7 V8 V9 V10 V11 V12
1: 2012 276 NA 0 S1 001 1 9218868437227407266 724135215 1590915056 NA 9.218868e+18
2: 2012 276 2 8 S1 001 1 9218868437227407266 9218868437227407266 154598 0 9.218868e+18
3: 2012 276 2 12 S1 001 1 9218868437227407266 5118863 21819477 NA 9.218868e+18
4: 2012 276 2 0 S1 011 8 3127133583 3127133583 9003982501 0 3.127134e+09
我錯過了什麼約integer64
,或者這是一個錯誤?如上所述,我可以繞過使用integer64="double"
,可能會失去一些精度,如幫助文件中所述。但意想不到的行爲是默認的integer64
...
這是在一臺運行Revolution R 3.0.2的Windows 8.1 64位機器以及運行kubuntu 13.10,CRAN-R 3.0.2的虛擬機上完成的。使用來自CRAN的最新穩定data.table(截至2014年2月7日的1.8.10)和1.8.11(1110,2014-02-04 02:43:19的修訂版,從zip作爲r-forge手動安裝建立被打破)在Windows上,只有在Linux上穩定的1.8.10。 bit64在兩臺機器上安裝並加載。
> sessionInfo()
R version 3.0.2 (2013-09-25)
Platform: x86_64-w64-mingw32/x64 (64-bit)
locale:
[1] LC_COLLATE=English_United States.1252 LC_CTYPE=English_United States.1252 LC_MONETARY=English_United States.1252 LC_NUMERIC=C
[5] LC_TIME=English_United States.1252
attached base packages:
[1] grid stats graphics grDevices utils datasets methods base
other attached packages:
[1] bit64_0.9-3 bit_1.1-11 gdata_2.13.2 xts_0.9-7 zoo_1.7-10 nlme_3.1-113 hexbin_1.26.3 lattice_0.20-24 ggplot2_0.9.3.1
[10] plyr_1.8 reshape2_1.2.2 data.table_1.8.11 Revobase_7.0.0 RevoMods_7.0.0 RevoScaleR_7.0.0
loaded via a namespace (and not attached):
[1] codetools_0.2-8 colorspace_1.2-4 dichromat_2.0-0 digest_0.6.4 foreach_1.4.1 gtable_0.1.2 gtools_3.2.1 iterators_1.0.6
[9] labeling_0.2 MASS_7.3-29 munsell_0.4.2 proto_0.3-10 RColorBrewer_1.0-5 reshape_0.8.4 scales_0.2.3 stringr_0.6.2
[17] tools_3.0.2
從幫助頁面,「此功能仍在開發中」。所以我期望作者將這個問題作爲一個錯誤 –
失望,這還沒有得到解決;如果有任何缺失值,它會使bit64軟件包對data.tables無用。我認爲這個問題必須用'fread',因爲我找不到任何方法來強制bit64包生成該值。它具有完全有效的NA值; 'as.integer64(NA)#' –
ClaytonJY
查看相關錯誤,https://github.com/Rdatatable/data.table/issues/488 –