I should add that the biggest problem is that Maple TA grades the right answer as wrong. I'm not sure how it obtains the other answer, so I'm not sure it that would be marked correct, or if it is expecting two answers. I probably would not have really looked at it if it accepted either answer. I'm not sure how likely it is that a student would stumble across Maple's "wrong" answer, so I don't know how big a problem that would be.

On a somewhat related note, if I feed it a sample of numbers with more than one mode, Maple's Mode function only reports the smallest one, and Maple TA will only accept that as the right answer. In that case I would like it to accept all of them as the only answer worth 100%.

TIA

Stephen

Thanks for the help. I changed to defining numeric values first, and cleaned my formulas up as suggested, and now it works the same in Maple and Maple TA. However, it doesn't work properly. An example:

A:=Array([[2.5,0],[4.5,3],[6.5,12],[8.5,25],[10.5,40],[12.5,47]]);

These (x,y) pairs are monotonically increasing.

evalf(solve(CurveFitting[Spline](A, x, degree = 1) = (47/2),x));

This finds the point on the spline connecting the points that marks the median. Maple gives two answers, however:

8.269230769, 8.300000000

When I check with my calculator, the first number gives 23.5 when substituted, but the second gives 23.7. When I perform a similar to find the first quartile (solve...=47/4,x) Maple (and Maple TA) gives three answers, the right one and two others.

If I plot the spline it is monotonically increasing, so the inverse shold be well defined. What don't I understand?

Thanks in advance.

Stephen