Thanks Preben, this works very well.

Thanks Preben, this works very well.

Thanks Robert and Duncan!

That wouldn't work as it would require me to type in all the 2nd column values, and there's thousands of them, and they change with each run!

That wouldn't work as it would require me to type in all the 2nd column values, and there's thousands of them, and they change with each run!

I have a table where the left key is a letter combination. The right key is the frequency of the letter combo. I want to graphically represent the frequency of each letter combo. Hence just plotting the second column should be enough, unless I can make the X access have coordinates my 1st column.

Again for illustrative purposes:

a   12

th   15

ch = 14

l = 12

k = 4

ck = 22

If I can just get a plot of the second column points I'm happy. For example I could say a=1, th = 2, ch = 3 etc, then get the ordered pairs:

(1,12), (2,15), (3,14), (4,12) etc, and just plot those. It would be nicer to be able to plot the original data somehow. Histogram maybe?

I apologize I wasn't clearer in my explanation. I'm multiplying these matrices to a high power, but I'm doing multiple runs/simulations. It's the runs I want to parallelize, not the multiplication.

The objects in the matrices are polynomials from a group ring, highly non-trivial. Ie their abstract variables (elements from a group), with a predeterined multiplication and I'm forming formal sums of them.

I apologize I wasn't clearer in my explanation. I'm multiplying these matrices to a high power, but I'm doing multiple runs/simulations. It's the runs I want to parallelize, not the multiplication.

The objects in the matrices are polynomials from a group ring, highly non-trivial. Ie their abstract variables (elements from a group), with a predeterined multiplication and I'm forming formal sums of them.