[GRASSLIST:211] Re: trouble using v.in.ascii

[Patton, Eric]

Vishal,

If you open your dbf in OpenOffice Calc, you are able to edit the data types
for each column. Here's an example from one of my dbf's:

CAT,N,11,0	CLASS_CODE,N,20,6	NEW_CLASS,C,50	Length,N,20,2
9	1	Rock Shores	961.07
178	1	Rock Shores	661.39
494	1	Rock Shores	768.71
508	1	Rock Shores	543.68
511	1	Rock Shores	989.52

The columns aren't going to line up nice in this email, but you get the
idea. Under the first column heading, 'CAT', there is an 'N' following it
signifying that this is a number, that 11 digit maximum is acceptable, and
that 0 decimals are allowed. db.columns in Grass would report that this is
an INTEGER datatype. You can edit these values to accept larger integer
values. If you want floating point datatype for this column, change the
final zero to match whatever the desired precision of your data.

The fourth column of data in my example corresponds to the Length column,
and is floating point data as you can see. If I wanted more decimal places,
I would chage the 2 to 6, say. The Length column could also be made into an
integer column simply by changing the ending '2' to a 0.


Hope this helps,

~ Eric.








-----Original Message-----
From: owner-GRASSLIST at baylor.edu
To: David Finlayson
Cc: Stefan Istvan; GRASS Users list
Sent: 3/17/2006 3:00 AM
Subject: [GRASSLIST:199] Re: trouble using v.in.ascii

Hi all,

ok i was able to get the points vector file done- silly syntax mistake
on my part-sorry.
But i could use some input on my attempts to connect my data dbf file to
the vector points file as layer 2.

What i did is, that i prepared a dbf file in Windows, then brought it
into my /dbf folder in Linux/Grass. Now this file is recognized when i
do db.describe(below):

GRASS 6.0.1 (ncdc):~/Desktop/NCDC > db.describe -c Data
ncols:11
Column 1: WMOID
Column 2: CITY
Column 3: YEAR
Column 4: YDAY
Column 5: TMAX
Column 6: TMIN
Column 7: TDAY
Column 8: RAIN
Column 9: VPD
Column 10: SRAD
Column 11: DAYLEN

BUT, the column 'WMOID' which i want to use as key, has suddenly become
float (it was integer when i made it in windows), and in fact  all
numeric columns are assigned double precision now when they were not
originally. How do i rectify this, since to identify a key tolink my
vector points to this data file, the key column should be integer?

I dont know postgresql, mysql etc, therefore i'm trying things this
way..

thanks,
vishal


On 3/17/06, David Finlayson < david.p.finlayson at gmail.com
<mailto:david.p.finlayson at gmail.com> > wrote:

Breezy also has the "flip" command (search synaptic) for converting
line endings.

-David

On 3/16/06, Stefan Istvan < stefi at geohidroterv.hu
<mailto:stefi at geohidroterv.hu> > wrote: 
> On Fri, 2006-03-17 at 11:30 +0530, Vishal Mehta wrote:
> > Hi all,
> >
> > I'm trying to bring in a comma separated text file as a point vector
> > file in GRASS using v.in.ascii , before i try to link it to a large
> > 33Mb dbf data file.
> >
> > The csv file looks like this with 39 rows and i removed the header
> > because grass60 (on Ubuntu breezy) does not have the skip option in 
> > the v.in.ascii command: the columns are WMOID, cityname, Latitude
and
> > Longitude. I want to use the WMOID as 'cat' in order to link my dbf
> > data later.
> >
> > 420710,AMRITSAR, 31.63,74.86
> > 421010,PATIALA,30.33,76.46
> > 421310,HISSAR,29.16,75.73
> > 421820,DELHI,28.58,77.20
> > 423390,JODHPUR,26.30,73.01
> > 423480,JAIPUR,26.81,75.80
> > 423690,LUCKNOW, 26.75,80.88
> > 424100,GAUHATI,26.10,91.58
> >
> > I Tried the following command and got the error:
> >
> >  cat reformat/STATIONS2.dat | v.in.ascii output=stations2 fs=','
> > columns='cat int, name(varchar(30), y double, x double' x=4 y=3
cat=1
> > WARNING: DOS text format found, attempting import anyway
> > Maximum input row length: 39
> > Maximum number of columns: 4 
> > Minimum number of columns: 4
> > column: 1  type: integer
> > column: 2  type: string  length: 18
> > column: 3  type: double
> > column: 4  type: double
> > DBMI-DBF driver error: 
> > SQL parser error in statement:
> > create table stations2 ( cat int, name(varchar(30), x double, y
> > double )
> > Error in db_execute_immediate()
> >
> > ERROR: Cannot create table: create table stations2 ( cat int, 
> >        name(varchar(30), x double, y double )
> >
> > I also get the same error with
> >
> > v.in.ascii input=reformat/STATIONS.csv output=gstations fs=','
skip=1
> > columns='cat int, label(varchar(30), y double, x double' x=4 y=3
cat=1 
> >
> There is an additional "(" sign in your columns definition. You should
> write columns='cat int, label varchar(30), y double, x double'.
> The other problem maybe that your file is in MSDOS format, but it
could 
> be that it doesn't bother Grass. Try to use dos2unix command to
convert
> it to unix format.
>
> Regards,
> Istvan
>
>


--
David Finlayson