Re: [Scilab-users] read format

[grivet]

Your suggestion does not work for me. Here is what I get:
-->M = csvRead('dataN.txt', ';', "double");
 -->M(1:10,:)
 ans  =
 Nan5.491.
Nan5.491.
Nan5.481.
Nan5.481.
Nan5.481.
which seems to check with the help for csvRead:

When the input argument "conversion" is equal to "double", the 
non-numeric fields within the .csv (e.g. strings) are converted into NaN.


Thank you for your time and help.
JP Grivet

Hello,

Well in fact the `evstr` call is not needed if you parse the csv and
interpreting directly the values as "double" using :

M = csvRead('/tmp/sample.csv', ';', "double");
values = M(:,2);

Regards,

--
Clément

Le lundi 12 octobre 2015 à 18:00 +0200, grivet a écrit :

Thank you Samuel, this works like a charm, even though the last step
(evstr) takes a couple of minutes (8 lines!).
For a completely different solution: read the file into emacs, define
a
rectangle of width 20 chars and height the whole file, delete this
rectangle, search for ".1" and erase every occurence of this string.
Thanks to everybody,
JP Grivet


Le 12/10/2015 14:26, grivet a écrit :

Hello,
I have a data file of about 80k lines. A typical line looks like
this:
 01/03/2015 00:01:00;5.49;1
(date time; value;parameter). I am only interested in the field
"value", which can be 3 or 4
characters wide (i.e. 5.49 or 5.4).
How can I extract the desired data from this file ?

.
Likely with
M  = csvRead(TheFileName, ";", ".", "string");
values = evstr(M(:,2));

Samuel Gougeon


___
users mailing list
users@lists.scilab.org
http://lists.scilab.org/mailman/listinfo/users


___
users mailing list
users@lists.scilab.org
http://lists.scilab.org/mailman/listinfo/users

Re: [Scilab-users] read format

[CHEZE David 227480]

But you said you were just interested in the last two numeric values : with 
double conversion into csvRead call it's the fastest way to get the relevant 
information and process it further. Would you need to get the timestamps data, 
you may call csvRead with "string" conversion and process the first column of 
the matrix of string to retrieve numeric value fields. 

Cheers,

David Chèze  

-Message d'origine-
De : users [mailto:users-boun...@lists.scilab.org] De la part de grivet
Envoyé : mardi 13 octobre 2015 12:09
À : Clément David <clement.da...@scilab-enterprises.com>; Users mailing list 
for Scilab <users@lists.scilab.org>
Objet : Re: [Scilab-users] read format

Your suggestion does not work for me. Here is what I get:
-->M = csvRead('dataN.txt', ';', "double");
  -->M(1:10,:)
  ans  =
  Nan5.491.
 Nan5.491.
 Nan5.481.
 Nan5.481.
 Nan5.481.
which seems to check with the help for csvRead:

When the input argument "conversion" is equal to "double", the non-numeric 
fields within the .csv (e.g. strings) are converted into NaN.

Thank you for your time and help.
JP Grivet
> Hello,
>
> Well in fact the `evstr` call is not needed if you parse the csv and 
> interpreting directly the values as "double" using :
>
> M = csvRead('/tmp/sample.csv', ';', "double"); values = M(:,2);
>
> Regards,
>
> --
> Clément
>
> Le lundi 12 octobre 2015 à 18:00 +0200, grivet a écrit :
>> Thank you Samuel, this works like a charm, even though the last step
>> (evstr) takes a couple of minutes (8 lines!).
>> For a completely different solution: read the file into emacs, define 
>> a rectangle of width 20 chars and height the whole file, delete this 
>> rectangle, search for ".1" and erase every occurence of this string.
>> Thanks to everybody,
>> JP Grivet
>>
>>> Le 12/10/2015 14:26, grivet a écrit :
>>>> Hello,
>>>> I have a data file of about 80k lines. A typical line looks like
>>>> this:
>>>>  01/03/2015 00:01:00;5.49;1 (date time; value;parameter). I 
>>>> am only interested in the field "value", which can be 3 or 4 
>>>> characters wide (i.e. 5.49 or 5.4).
>>>> How can I extract the desired data from this file ?
>>> .
>>> Likely with
>>> M  = csvRead(TheFileName, ";", ".", "string"); values = 
>>> evstr(M(:,2));
>>>
>>> Samuel Gougeon
>>>
>> ___
>> users mailing list
>> users@lists.scilab.org
>> http://lists.scilab.org/mailman/listinfo/users

___
users mailing list
users@lists.scilab.org
http://lists.scilab.org/mailman/listinfo/users
___
users mailing list
users@lists.scilab.org
http://lists.scilab.org/mailman/listinfo/users

Re: [Scilab-users] read format

[sgougeon]

Hello,

- Mail original -
>De: "CHEZE David 227480" 
>
>.../... Would you need to get the timestamps data, you may call csvRead with 
>"string" conversion and process the first column of the matrix of string to 
>retrieve numeric value fields. 

For this part, using mfscanf() as suggested by Serge will be more 
straightforward, instead of csvRead().

Samuel
___
users mailing list
users@lists.scilab.org
http://lists.scilab.org/mailman/listinfo/users

Re: [Scilab-users] read format

[Serge Steer]

It is possible to achieve what you want using the following instruction.
It is a little tricky, may be a C format expert can do it in a more
simple way.
u=mopen("",'r')
x=mfscanf(-1,u,"%*s %*2s%*[:]%*2s%*[:]%*2s%*[;]%g%*s");
mclose(u)

Serge Steer
Le 13/10/2015 16:19, sgoug...@free.fr a écrit :
> Hello,
>
> - Mail original -
>> De: "CHEZE David 227480" 
>>
>> .../... Would you need to get the timestamps data, you may call csvRead with 
>> "string" conversion and process the first column of the matrix of string to 
>> retrieve numeric value fields. 
> For this part, using mfscanf() as suggested by Serge will be more 
> straightforward, instead of csvRead().
>
> Samuel
> ___
> users mailing list
> users@lists.scilab.org
> http://lists.scilab.org/mailman/listinfo/users
>

___
users mailing list
users@lists.scilab.org
http://lists.scilab.org/mailman/listinfo/users

Re: [Scilab-users] read format

[grivet]

Thank you Samuel, this works like a charm, even though the last step 
(evstr) takes a couple of minutes (8 lines!).
For a completely different solution: read the file into emacs, define a 
rectangle of width 20 chars and height the whole file, delete this 
rectangle, search for ".1" and erase every occurence of this string.

Thanks to everybody,
JP Grivet


Le 12/10/2015 14:26, grivet a écrit :

Hello,
I have a data file of about 80k lines. A typical line looks like this:
01/03/2015 00:01:00;5.49;1
(date time; value;parameter). I am only interested in the field 
"value", which can be 3 or 4

characters wide (i.e. 5.49 or 5.4).
How can I extract the desired data from this file ?

.
Likely with
M  = csvRead(TheFileName, ";", ".", "string");
values = evstr(M(:,2));

Samuel Gougeon



___
users mailing list
users@lists.scilab.org
http://lists.scilab.org/mailman/listinfo/users

Re: [Scilab-users] read format

[Serge Steer]

with s given as follow
s=['01/03/2015 00:01:00;5.49;1'
'01/03/2015 00:01:01;5.9;1'
'01/03/2015 00:01:11;4.9;1']
The following instructions returns the expexted results
msscanf(-1,s,"%*s %*2s%*[:]%*2s%*[:]%*2s%*[;]%g[^;]%*[;]")

But I do not succeed if the data are stored in a file
u=mopen("","r")
mfscanf(-1,u,"%*s %*2s%*[:]%*2s%*[:]%*2s%*[;]%g[^;]%*[;]")
returns an error

I think a C format specialist could be able to find my error

Serge
Le 12/10/2015 14:26, grivet a écrit :
> Hello,
> I have a data file of about 80k lines. A typical line looks like this:
> 01/03/2015 00:01:00;5.49;1
> (date time; value;parameter). I am only interested in the field
> "value", which can be 3 or 4
> characters wide (i.e. 5.49 or 5.4).
> How can I extract the desired data from this file ?
> Thank you for your help.
> JP Grivet
> ___
> users mailing list
> users@lists.scilab.org
> http://lists.scilab.org/mailman/listinfo/users
>

___
users mailing list
users@lists.scilab.org
http://lists.scilab.org/mailman/listinfo/users

Re: [Scilab-users] read format

[Clément David]

Hello,

Well in fact the `evstr` call is not needed if you parse the csv and
interpreting directly the values as "double" using :

M = csvRead('/tmp/sample.csv', ';', "double");
values = M(:,2);

Regards,

--
Clément

Le lundi 12 octobre 2015 à 18:00 +0200, grivet a écrit :
> Thank you Samuel, this works like a charm, even though the last step 
> (evstr) takes a couple of minutes (8 lines!).
> For a completely different solution: read the file into emacs, define
> a 
> rectangle of width 20 chars and height the whole file, delete this 
> rectangle, search for ".1" and erase every occurence of this string.
> Thanks to everybody,
> JP Grivet
> 
> > Le 12/10/2015 14:26, grivet a écrit :
> > > Hello,
> > > I have a data file of about 80k lines. A typical line looks like
> > > this:
> > > 01/03/2015 00:01:00;5.49;1
> > > (date time; value;parameter). I am only interested in the field 
> > > "value", which can be 3 or 4
> > > characters wide (i.e. 5.49 or 5.4).
> > > How can I extract the desired data from this file ?
> > .
> > Likely with
> > M  = csvRead(TheFileName, ";", ".", "string");
> > values = evstr(M(:,2));
> > 
> > Samuel Gougeon
> > 
> 
> ___
> users mailing list
> users@lists.scilab.org
> http://lists.scilab.org/mailman/listinfo/users
___
users mailing list
users@lists.scilab.org
http://lists.scilab.org/mailman/listinfo/users

[Scilab-users] read format

[grivet]

Hello,
I have a data file of about 80k lines. A typical line looks like this:
01/03/2015 00:01:00;5.49;1
(date time; value;parameter). I am only interested in the field "value", 
which can be 3 or 4

characters wide (i.e. 5.49 or 5.4).
How can I extract the desired data from this file ?
Thank you for your help.
JP Grivet
___
users mailing list
users@lists.scilab.org
http://lists.scilab.org/mailman/listinfo/users

Re: [Scilab-users] read format

[Samuel Gougeon]

Hello,

Le 12/10/2015 14:26, grivet a écrit :

Hello,
I have a data file of about 80k lines. A typical line looks like this:
01/03/2015 00:01:00;5.49;1
(date time; value;parameter). I am only interested in the field 
"value", which can be 3 or 4

characters wide (i.e. 5.49 or 5.4).
How can I extract the desired data from this file ?

.
Likely with
M  = csvRead(TheFileName, ";", ".", "string");
values = evstr(M(:,2));

Samuel Gougeon


___
users mailing list
users@lists.scilab.org
http://lists.scilab.org/mailman/listinfo/users

Re: [Scilab-users] read format

[Collewet Guylaine]

Hello,

Perhaps this could help (there is probably a more direct way to do it)

str1="00:01:00;5.49;1";
str2=strchr(str1,';');   //str2= ";5.49;1"
str3=strtok(str2,';'); // str3 = "5.49"
d =  strtod(str3);

Regards

Guylaine

-Message d'origine-
De : users [mailto:users-boun...@lists.scilab.org] De la part de grivet
Envoyé : lundi 12 octobre 2015 14:27
À : users@lists.scilab.org
Objet : [Scilab-users] read format

Hello,
I have a data file of about 80k lines. A typical line looks like this:
 01/03/2015 00:01:00;5.49;1
(date time; value;parameter). I am only interested in the field "value",
which can be 3 or 4 characters wide (i.e. 5.49 or 5.4).
How can I extract the desired data from this file ?
Thank you for your help.
JP Grivet
___
users mailing list
users@lists.scilab.org
http://lists.scilab.org/mailman/listinfo/users
___
users mailing list
users@lists.scilab.org
http://lists.scilab.org/mailman/listinfo/users

Re: [Scilab-users] read format

[Samuel Gougeon]

Le 12/10/2015 18:07, Clément David a écrit :

M = csvRead('/tmp/sample.csv', ';', "double");
values = M(:,2);

Indeed, or simply

M = csvRead(theFileName, ';');

Without trying, i thought that meeting the date "01/03/2015  00:01:00;" would 
yield an error. But it is managed softly: it yields %nan instead.
Scilab is great when it manages things softly :)
This is why ieee(2) would be a soft default ;)

Samuel

___
users mailing list
users@lists.scilab.org
http://lists.scilab.org/mailman/listinfo/users