This here looks even better
grid >TAB cut &.> (<;._2) fd6
2012/11/11 Björn Helgason
> I copied the contents of this script into the scratch window of JHS and
> ran it and the grid works just fine there as well as in GTK
>
> fd=: 0 : 0
> NB. a \t 1 \t 2 \t 3
> NB. b \t \t 5 \t 6
> NB. c \t 7 \t
I copied the contents of this script into the scratch window of JHS and ran
it and the grid works just fine there as well as in GTK
fd=: 0 : 0
NB. a \t 1 \t 2 \t 3
NB. b \t \t 5 \t 6
NB. c \t 7 \t \t 9
NB. d \t 10 \t 11 \t 12
NB. \t 13 \t 14 \t 15
)
[fd2=:fd rplc '\t \t';'\t NA \t'
[fd3=:fd2 rpl
fd=:'mytestfile.txt' NB. TAB text file
NB. a \t 1 \t 2 \t 3
NB. b \t \t 5 \t 6
NB. c \t 7 \t \t 9
NB. d \t 10 \t 11 \t 12
NB. \t 13 \t 14 \t 15
1a 1 2 3
2b NA 5 6
3c 7 NA 9
4d 10 11 12
5 13 14 15
fd=: 0 : 0
NB. a \t 1 \t 2 \t 3
NB. b \t \t 5 \t 6
NB. c \t 7 \t \t 9
NB. d
Thanks for everything.
The utility function readdsv is functionally enough for my purpose. But I
noticed that high functionality results in less speed. It takes much time for
reading more than 400 million lines with each line of a hundred variables.
Thanks again.
Toshinari
On 2012/11/10,
Note that "readdsv" takes an optional left argument to specify the
delimiter and quotes character, with a default of (TAB;'"'), e.g.
readdsv 'testTAB.txt'
++-+-+
|This is entry |1|... more stuff |
++-+-+
You might want to look at the 'tables/dsv' and 'tables/csv' addons.
http://www.jsoftware.com/jwiki/Addons/tables/dsv
http://www.jsoftware.com/jwiki/Addons/tables/csv
The 'tables/csv' addon is basically a special case of the more general
'tables/dsv', where the delimiter is set to ',' and the quote
We did a "Beginner's Regatta" at NYCJUG earlier this year in which we
explored importing tabular data into J from delimited files:
http://www.jsoftware.com/jwiki/NYCJUG/2012-05-08#Beginner.27s_regatta with
a follow-up the following month:
http://www.jsoftware.com/jwiki/NYCJUG/2012-06-12#Follow-up_t
Purely a convenience thing, but using the verb freads rather than fread
will take care of the line-ending issue, by converting to the J standard of
LF-delimited lines. The verb fwrites will convert J line-endings to the
conventional line-ending for your current platform.
On Nov 8, 2012 12:34 PM, "R
(responding inline because that makes sense to me, here.)
On Wed, Nov 7, 2012 at 11:53 PM, Toshinari Kamakura
wrote:
> Thank you for quick responding to my mail.
That was pure luck -- I just happened to be checking my email shortly
after your previous post arrived.
> My desire is just to handle
J has no explicit NA value. It is not simple to produce results like
you want because even with all numeric data that include missing
values, the missing values of "NA" are not numeric and complicate the
production of results. Numeric NA values can be represented with J _.
but this is very tricky a
Thank you for quick responding to my mail.
My desire is just to handle the missing values as R language can do.
> u=read.table("mytestfile.txt",header=F,na.strings="",sep="\t")
> u
V1 V2 V3 V4
1a 1 2 3
2b NA 5 6
3c 7 NA 9
4d 10 11 12
5 13 14 15
The R system inputs NA v
I am not sure what your line ends look like.
A general implementation might be:
<;._2@,&TAB;._2 -.&CR ,&LF fread 'mytestfile.txt'
This will have an extra blank row at the bottom if your file is
newline terminated, in which case you might want:
<;._2@,&TAB;._2 -.&CR fread 'mytestfile.txt'
Hi
I would like to know J's manipulation for reading text file.
fd=:'mytestfile.txt' NB. TAB text file
NB. a \t 1 \t 2 \t 3
NB. b \t \t 5 \t 6
NB. c \t 7 \t \t 9
NB. d \t 10 \t 11 \t 12
NB. \t 13 \t 14 \t 15
fd2=:'mytestfile.csv' NB. CSV test file
load'files misc'
freads fd2
a,1,2,3
b,,5,6
c
13 matches
Mail list logo
