Strange output using symbols in a matrix

I always seem to have some weird results when using symbols in matrices. I wrote a simple function that generates a transformation matrix based on arguments.

    function rot = dh_t (a,alpha, d, theta)
       rot = [cos(theta)            -sin(theta)           0           a;...
              sin(theta)*cos(alpha) cos(theta)*cos(alpha) -sin(alpha) -d*sin(alpha);...
              sin(theta)*sin(alpha) cos(theta)*sin(alpha) cos(alpha)  d*cos(alpha);...
              0                     0                     0           1]
    end

When I call this function:

    dh_t(0,pi/2,L1,t2)

I get an ugly output.

   [                                                     cos(t2),                                                    -sin(t2),                                                 0,                                                      0]
   [ (4967757600021511*sin(t2))/81129638414606681695789005144064, (4967757600021511*cos(t2))/81129638414606681695789005144064,                                                -1,                                                    -L1]
   [                                                     sin(t2),                                                     cos(t2), 4967757600021511/81129638414606681695789005144064, (4967757600021511*L1)/81129638414606681695789005144064]
   [                                                           0,                                                           0,                                                 0,                                                      1]

I stick this in my calculator where I wrote an equivalent function:

    [cos(t2)   −sin(t2)   0  0]
    [sin(t2)   cos(t2)    0  0]
    [0         0          1 l1]
    [0         0          0  1]

Suggestions are welcome on how to handle this. Thanks!

[madhan ravi]

I stick this in my calculator where I wrote an equivalent function:

What do you mean by this? can you elaborate more?

[Jason Gauthier]

Yes. It doesn't happen with my Ti-Inspire calculator.

Answer 1

[Walter Roberson]

dh_t(0,sym(pi)/2,L1,t2)

[Jason Gauthier]

oh, that's very nice.

Answer 2

[Star Strider]

Use the vpa (link) function.

Result = vpa(dh_t(0,pi/2,L1,t2), 3)

[Jason Gauthier]

[          cos(t2),     -1.0*sin(t2),        0,           0]
[ 6.12e-17*sin(t2), 6.12e-17*cos(t2),     -1.0,     -1.0*L1]
[          sin(t2),          cos(t2), 6.12e-17, 6.12e-17*L1]
[                0,                0,        0,         1.0]

I don't think so. That doesn't provide clarity.

[Jason Gauthier]

although that appears more to be an issue with cos/sin inaccuracies.

[Star Strider]

I suspect that the TI-inspire is not using the extended-precision representation that the Symbolic Math Toolbox does. It may instead be using the IEEE 754 standard, and considers 6.12e-17 as essentially 0.