How to simplify symbolic calculation

Hi there everybody,

I have been trying to do symbolic calculation on Matlab, but the solution gives very unnecessary number which is described in below: "6.123e-17*qdot1 + 1.0*qdot2 + 1.0*qdot3 + qdot5 - 6.123e-17*cos(q5)*(3.749e-33*qdot1..."

This is an answer matrix column. I'd like Matlab to calculate 6.123e-17*qdot as 0 and like to avoid 1.0*qdot3. Could you help me about this problem?

Thanks in advance.

Have a nice day.

Best regards, Cem ÇAKMAK

Answer 1

[madhan ravi]

see https://in.mathworks.com/matlabcentral/answers/168162-how-can-i-round-the-coefficient-of-a-term-containing-symbolic-variables#answer_163356 similar example

[Walter Roberson]

I recall that there is another similar discussion that I was one of the participants of. I am away from my computer so I cannot locate it at the moment.

[madhan ravi]

ok sir Walter will search through

[madhan ravi]

see https://in.mathworks.com/matlabcentral/answers/379115-why-matlab-make-a-mistake-in-calculating-sin-and-cos-while-teta-in-wiling-to-zero#answer_301889 and https://ww2.mathworks.cn/matlabcentral/answers/66952-numerical-round-off-instability-help-how-to-get-more-precision#answer_78415

[Walter Roberson]

Those were not the ones I was thinking of; the one I was thinking of was quite similar to the current question.

Unfortunately I cannot locate it at this time.

The general idea is that it is possible to use evalin(symengine) or feval(symengine) to run some MuPAD code which will examine the structure of the expressions, locate constants with sufficiently small absolute value, and substitute zeros for them. This would probably involve using the MuPAD match() and subsex() and a loop.

John's answer that you linked to is still important: for sufficiently large numeric replacement values for symbolic variables, neglecting "small" contributions becomes risky.