| T O P I C R E V I E W |
| antekb |
Posted - 02/01/2007 : 10:31:45 AM
Origin Version :7.0 SR 2
Operating System:Windows XP
Hi
Is there a method to copy rows from 1 to 40 and put it into a new column?
|
| 10 L A T E S T R E P L I E S (Newest First) |
| Mike Buess |
Posted - 02/07/2007 : 12:34:00 PM
But ds is still a dataset and you can operate on the whole dataset with a scalar.
Worksheet wksNormalized;
wksNormalized.Create("Origin.otw");
for(i=0; i<wksNormalized.GetNumCols(); i++) // delete existing columns
{
wksNormalized.DeleteCol(0);
}
Dataset ds, dsData; // Mz-tau-t, normalized
double dMO = wks.Cell(0,0); // dMO is value at first row of first column
for( int c = 2; c < (b+1); c++)
{
int iCol = wksNormalized.AddCol();
ds.Attach(wks,c);
dsData.Attach(wksNormalized,iCol);
dsData = (dMO - ds) / (2.0 * dMO); // vector math
}
}
Mike Buess
Origin WebRing Member |
| antekb |
Posted - 02/07/2007 : 08:18:06 AM
Hi Mike
Originally my data is a 2D array [t x tau] where t=40, tau=vary.It is
stored as single column (first few rows are labels-ASCII import procedure
just make them column label-they do not interrupt, next tau (b in script
code) rows are the values of time spacings between measurements; after that
there are t=40 rows containing time base of a single measurement.Last tau x
t(=40) rows contain amplitudes of all measurements.
I am sorry for this, but...
The problem I have just found is that normalization procedure requires that
the values of all rows from (b-1) columns should be calculated as follows:
((M0-Mz(t,tau))/2*M0) where M0 is a value stored in first row of first column->after part one. M0 is NOT a whole column, SORRY (Mea culpa..) !!
Worksheet wksNormalized;
wksNormalized.Create("Origin.otw");
for(i=0; i<wksNormalized.GetNumCols(); i++) // delete existing columns
{
wksNormalized.DeleteCol(0);
}
Dataset ds2(wks,1); // MO= this should be single value instead of column so I'm afraid that vector math won't work here...How to refer to this value?
Dataset ds, dsData; // Mz-tau-t, normalized
for( int c = 2; c < (b+1); c++)
{
int iCol = wksNormalized.AddCol();
ds.Attach(wks,c);
dsData.Attach(wksNormalized,iCol);
dsData = (ds2 - ds) / (2*ds2); // vector math
}
}
|
| Mike Buess |
Posted - 02/06/2007 : 10:06:05 AM
1) This will get the name of new wks but I don't think you need it.
string wksNormalizedName = wksNormalized.GetPage().GetName();
2) At this point none of your datasets are declared so your Part 2 code won't work. Also, you don't need to calculate the normalized dataset one row at a time. Vector math is much faster. Also2, it doesn't look like you have enough columns in your new wks? By now I'm confused about what a, b and c mean (doesn't a=1?) but you can try this for Part 2...
Worksheet wksNormalized;
wksNormalized.Create("Origin.otw");
for(int i=0; i<wksNormalized.GetNumCols(); i++) // delete existing columns
{
wksNormalized.DeleteCol(0);
}
Dataset ds2(wks,1); // MO?
Dataset ds, dsData; // Mztau?, normalized?
for( int c = 2; c < (b+1); c++)
{
int iCol = wksNormalized.AddCol();
ds.Attach(wks,c);
dsData.Attach(wksNormalized,iCol);
dsData = (ds2 - ds) / ds; // vector math
}
Mike Buess
Origin WebRing Member |
| antekb |
Posted - 02/06/2007 : 08:38:48 AM
Hi
This time I need to:
1) normalize the data, which are output from the last code: so I have 25 columns, 40 rows each
First column= original data, 2nd col = M0, remaining (a-1) columns is Mztau. To do normalization I need to calculate (M0-Mztau)/2*M0 for every (a-1) columns
2) put the result of this into new worksheet
I tried to use the following example
http://www.originlab.com/forum/topic.asp?TOPIC_ID=2378
So I got(of course it has no right to work!->only to show an idea):
[code]
void make(int b)
{
if(b < 1)
{
printf("incorrect column number specification!\n");
return;
}
Worksheet wks = Project.ActiveLayer();
int a = wks.GetNumCols();
for(int i=0; i<a; i++)
{
Dataset ds1(wks,i);
for(int k=b+40; k<(ds1.GetSize()); k+=40)
/* k=b+40 because these rows contains data I need later on */
{
int iNew = wks.AddCol();
Dataset ds2(wks,iNew);
Data_copy(&ds2,&ds1,k,k+39,0);
}
}
/* end of part one*/
////////////////////////////////////////////////////////////////////////////////////////
Worksheet wksNormalized;
wksNormalized.Create("Origin.otw")
/* (1) what would be the name of this worksheet?? */
for( int c = 2; c < (b+1); c++)
/*(c) should loop through columns that are to be normalised; c < (b+1) because first col is original (RAW) data, 2nd col contain M0 values (assuming RAW data col has index 0)*/
{
int iSize = ds2.GetSize();
/*(2)here I want to "get" lenght of dataset ds2(should be 40 rows) - but it seems to me that every (b)'s column should be a separate dataset...how to do that? */
for(int iRow = 0; iRow < iSize; iRow++)
{
dsData[iRow] = ( ds2[iRow]-ds[iRow]/(2*ds2[iRow]) ;
}
/*normalizedData(in new=normalizedWKS) =(M0-Mztau)/2*M0{from part one}*/
}
}
Q:
3)How to refer to columns in a different worksheet?
|
| antekb |
Posted - 02/06/2007 : 06:02:25 AM
Thanks Mike,
it works very well
I'm going to develop this further...
Best regards!! |
| Mike Buess |
Posted - 02/05/2007 : 09:40:24 AM
Hi Antek,
Try this...
void copy_col()
{
Worksheet wks = Project.ActiveLayer();
int a = wks.GetNumCols();
for(int i=0; i<a; i++)
{
Dataset ds1(wks,i);
for(int k=0; k<ds1.GetSize(); k+=40)
{
int iNew = wks.AddCol();
Dataset ds2(wks,iNew);
Data_copy(&ds2,&ds1,k,k+39,0);
}
}
}
Mike Buess
Origin WebRing Member |
| antekb |
Posted - 02/05/2007 : 06:16:05 AM
Hi Mike
Thanks for help
The code works, but after first (a) columns has been copied (and this should terminate the loop) program copy first 40 rows to column (a+1), then rows 41 to 80 to column a+25 and so on(as a result of this error I get 552 additional columns).
To be clear - source column in this notation would have (a-1)index.
With kind regards
Antek |
| Mike Buess |
Posted - 02/02/2007 : 08:18:46 AM
1. Your inner loop is wrong. You move rows 0-39, 1-40, 2-41, etc. rather than 0-39, 40-79, 80-119, etc.
2. nDestFrom=-1 means nDestFrom=nSrcFrom rather than nDestFrom=0.
3. You create and declare the new columns inside the wrong loop.
Try this...
for(int i=0; i<a; i++)
{
Dataset ds1(wks,i);
for(int k=0; k<a*40; k+=40)
{
int iNew = wks.AddCol();
Dataset ds2(wks,iNew);
Data_copy(&ds2,&ds1,k,k+39,0);
}
}
Mike Buess
Origin WebRing Member
Edited by - Mike Buess on 02/02/2007 08:50:33 AM |
| antekb |
Posted - 02/02/2007 : 05:18:48 AM
hi
thanks for help!!
but.. I need it more...!!!!
I tried to modify your code to move NEXT 40 rows into NEXT new column
(and repeat this "a" times)
void move(int a) // "a" is a number of column
{
Worksheet wks = Project.ActiveLayer();
for (int i=0;i < a ;i++)
{
int iNew = wks.AddCol();
Dataset ds1(wks,i);
Dataset ds2(wks,iNew);
for (int k=0; k < a*40;k++)
{
Data_copy(&ds2,&ds1,k,k+39,-1);//(-1) = because I want //to move(old)row 40 to row 0 in new column and so on
}
}
}
This code works well, except that it just copy everything "a" times.
To be precise: I try to make (a)columns, 40 rows each, from one column which is (a)*40 rows long
|
| Mike Buess |
Posted - 02/01/2007 : 11:52:21 AM
This copies rows 1-40 from first column to a new column...
Worksheet wks = Project.ActiveLayer();
int iNew = wks.AddCol();
Dataset ds1(wks,0);
Dataset ds2(wks,iNew);
Data_copy(&ds2,&ds1,0,39)
Mike Buess
Origin WebRing Member |
|
|
