ACCEPTED SOLUTION
Accepted Solutions
Your worksheet doesn't show your problem, but I'm guessing you're trying to assign units to alpha?
If so, then the problem is the string header.
Stuart
That was my initial attempt but I couldn't get the submatrix arguments right
To extract the numbers from matrix X, call the submatrix function. The arguments for submatrix are as follows: the name of the matrix, the indices of the first and of the last rows to extract, the indices of the first and of the last columns to extract
a: = submatrix (name of matrix, the indices of the first and last rows to extract, indices of the first and last columns to extract)
a = submatrix( x, row 1, to row last(of matrix X), column 0, to column 0 , column 0)
Thats because of different labels. The built in function is labelled "Function", while the variable is labelled "Variable".
They both can co-exist, but whenever you write "matrix" after you defined the variable, Prime assumes that you mean your variable and not the built-in function. You would have to re-label as "Function" to have access to the built-in function.
As this can be very cumbersome and confusing, I highly second your advice not to use names of built-in function as variable or user-defined function names. BTW, a user-defined function is also automatically labelled "Variable" ad so can also coexist with a built-in one.
So then why do I need to define the units as such?
I'm afraid this will create errors down the line.
without that definition, this is still after doing submatrix as you advised.
And for it to work I need to extract just the column from excel, then extract a submatrix from that same column?
I can't extract the matrix, and then a submatrix column from that matrix.
Thanks,
I can't extract the matrix, and then a submatrix column from that matrix.
I am not sure which problem you experience.
Like Stuart i cannot see the error ("when I enter unit for the vectors, it doesn't like it") you described in your initial posting.
In one of the screenshots you posted it can be seen m:=rows(matrix)=0 which usually means thats matrix is just a normal scalar variable. The matrix seems to be named "X" in he screen shot.
Your "submatrix" command fails because "last" can only be applied to vectors (n x 1 matrices). You would have to use "last(X^<0>)", where "X^<0>" should denote the first column of your matrix X.
If you experience still other problems please describe them in more detail.
I attach a sheet where I extracted the alpha column, stripped the header and added the unit centimeter.
Done on the vector alpha you already extracted and then directly from the matrix.
If you will never change the value of ORIGIN from 0 to something else, you sure can make the expression a bit shorter by directly using that number.
EDIT: From re-reading what you wrote I would guess that the faulty use of "last" with a matrix as argument was the reason for the confusion?
BTW, instead of last(matrix^<ORIGIN>) you could also use ORIGIN+rows(matrix)-1 for the very same result.
@cadtelsim wrote:
So then why do I need to define the units as such?
I'm afraid this will create errors down the line.
without that definition, this is still after doing submatrix as you advised.
Your Excel worksheet header gives the unit of alpha as "/cm", which I assume means 1/cm.
You probably already know this quite well, but ... If you were to multiply alpha by ft (feet) or ns (nanosecond), Mathcad would internally convert the quantity into metres and seconds and adjust the numerical value accordingly. When you evaluate the quantity it will display in SI units. Typing the correct unit in (totally replacing the initially displayed SI unit) changes the numerical value accordingly.
Mathcad treats 1/cm the same way. That is why after multiplying alpha by 1/cm, you see 1.73 x 10^8 1/m but 1.73 x 10^6 1/cm ... the 1/cm has rescaled the numerical value of the quantity, and it matches the numerical value shown in your Excel worksheet.
You need to multiply lambda by nm to get the correct units (as indicated by the column header)
So if there are problems further downstream then it means there is likely a problem with your equations/formulae/etc. A common cause of such problems is a constant that was derived using non-SI units and hasn't been recalculated for SI units, or not scaled properly for multiples of SI units (eg, an eqn assumes mm).
As Tetsuo points out, if *you* aren't too bothered about importing the header row, then you can specify the input range as "Sheet1!A2:C92". However, it's often useful to import it anyway, as it allows you or a reviewer to check that the correct units were applied to all input variables
Stuart
@cadtelsim wrote:
And for it to work I need to extract just the column from excel, then extract a submatrix from that same column?
I can't extract the matrix, and then a submatrix column from that matrix.
Thanks,
The need to extract or remove the first row of a matrix is such a common task that it is worthwhile writing functions to do just that and keeping them in your default template.
Stuart
The advantage is that I don't want to see this equation in Excel.
Is that considered compelling enough?
@cadtelsim wrote:
The advantage is that I don't want to see this equation in Excel.
Is that considered compelling enough?
Meh, personal preference. Seen worse. 🙂
(And debugged worse!)
But, getting back to the original question in context, how does using an Excel Component make it any easier to remove a row from a matrix the OP needs to import anyway (and then still having to add units to the resulting vector, anyway)? A simple submatrix would seem to do the job without invoking the Excel Monster's name and summoning it forth.
Stuart
Again - don't use inline evaluation.
Evaluate lambda in a separate region.
I think just using the submatrix command with the correct usage of "last" would be easier ...
This time I opened your sheet (which again did not show what your screenshot showed!!) with Prime 9
@cadtelsim wrote:
This I could do, but could you tell me why the unit doesn't work?
It did work. As I said in a previous message, Mathcad converts units to SI (base) units, eg metres, and displays them as such by default. You must multiply by the unit in which you want to see the result.
Stuart
