So, "solve" will always assume that the expression in each position of the vector is an equation,
or an inequality, so sometimes you have the choice between "assume" and an additional constraint in the vector (and I would normally prefer "assume").
and it will assume equal with 0 if not defined otherwise.
Its unclear to me why you want to do the things more complicated as necessary. Using the symbolic solve you won't need the "=<vector of 1's>" as already described and shown.
But the way you did it works, too.
It fails in your sheet because the variables x0, etc already are assigned values. Every symbolic solve will fail in that case.
So either you try it in a fresh worksheet without any calculations done before or you clear the variable x (consisting of x0, etc,) for symbolic evaluations by typing x:=x.
Werner Exinger wrote:
I use MC15, but my file should work in MC14 as well, unless there is a bug which was fixed in the meantime.
Which maintainance release of MC14 are you using?
Did you change anything in the file other than the assignment V:=
I using MC14 M020, Werner.
I think the problem is already discussed here and "solved" - its because the solveblock does not like the use of vetcor elements as solve variables (still not sure if its not a bug) and I think you have provided a workaround yourself by defining the vector beforhand using "normal" variables.
BTW, I used stack(), and aa, etc. instead of a correctly dimensioned vector and alpha, .. just out of lazyness 😉