What happened for "simple" command in matlab 2016

Hello

I already installed Matlab 2016a and I noticed that "simple" command doesn't exist in matlab anymore and I'm not sure when this command has been removed from matlab. Please note that "simplify" can not solve my current problem. I am found that "solve" command doesn't work as good as previous versions. For example it could not solve a sixth degree equation and the solution is based on the "root".

Does anybody know the alternative commands for these two.

Thanks

Answer 1

[Benjamin Kraus]

It looks like the simple command (from the Symbolic Math Toolbox) was removed in R2015a. It is present in R2014b. The R2014b doc page says simple "will be removed in a future release.". The R2015a doc page says simple "has been removed."

Both pages say to use simplify instead.

In the future, this type of information can often be found in the Release Notes (I prefer to click "expand all in page" and then use the search feature).

As for alternatives, you will have to provide more information (such as an example) about the problem you are trying to solve, and why simplify and solve are not doing what you want before people can suggest alternatives.

[hamed shorakaei]

Thank you for your answering.

I have a huge sentence based on the some parameters, I need to extract the coefficient of a parameter, so by using subs and zeroing the other parameters and then simple command I succeeded to extract my desired coefficient (coeffs command doesn't work correctly here). when I used simplify the coefficient went zero and gave me wrong answer. BTW, I think it need to back to matlab.

About solve, please see the following code:

[Walter Roberson]

There are no closed form solutions for most polynomials of degree 5 or higher.

Earlier versions of MATLAB would have expressed the same thing using RootOf() instead of root(), or would have given numeric approximations. If you want the numeric approximations, you can vpa() the current output. The symbolic placeholder is much more useful for algebraic computation such as not accidentally back-substituting the wrong root branch when doing further computations.

[hamed shorakaei]

Yes, Thank you. "vpa" works.

Answer 2

[Walter Roberson]

simple was removed in R2015a. https://www.mathworks.com/help/releases/R2015a/symbolic/release-notes.html#R2015a

Answer 3

[Lei LIU]

I meet the same problem, i.e., the result of 'simplify' is not desired but 'simple' works well.

Answer 4

[Deniz Muhacir]

Hi everyone!

Can anybody solve this issue? Has anyone got alterne code for simple? Because i need to make simple a matrix that :

(24678615572571482867467662723121*sin(theta1))/6582018229284824168619876730229402019930943462534319453394436096 + (4967757600021511*cos(theta3)(cos(theta1)*sin(theta2) + (4967757600021511*cos(theta2)*sin(theta1))/81129638414606681695789005144064))/81129638414606681695789005144064 + (4967757600021511*sin(theta3)(cos(theta1)cos(theta2) - (4967757600021511*sin(theta1)*sin(theta2))/81129638414606681695789005144064))/81129638414606681695789005144064 + cos(theta4)((4967757600021511*cos(theta3)(cos(theta1)*sin(theta2) + (4967757600021511*cos(theta2)*sin(theta1))/81129638414606681695789005144064))/81129638414606681695789005144064 - sin(theta1) + (4967757600021511*sin(theta3)(cos(theta1)cos(theta2) - (4967757600021511*sin(theta1)*sin(theta2))/81129638414606681695789005144064))/81129638414606681695789005144064) + sin(theta4)(cos(theta3)(cos(theta1)*cos(theta2) - (4967757600021511*sin(theta1)*sin(theta2))/81129638414606681695789005144064) - sin(theta3)(cos(theta1)*sin(theta2) + (4967757600021511*cos(theta2)*sin(theta1))/81129638414606681695789005144064))

(4967757600021511*sin(theta3)((4967757600021511*cos(theta1)*sin(theta2))/81129638414606681695789005144064 + cos(theta2)*sin(theta1)))/81129638414606681695789005144064 - (4967757600021511*cos(theta3)((4967757600021511*cos(theta1)cos(theta2))/81129638414606681695789005144064 - sin(theta1)*sin(theta2)))/81129638414606681695789005144064 - (24678615572571482867467662723121*cos(theta1))/6582018229284824168619876730229402019930943462534319453394436096 + sin(theta4)(cos(theta3)((4967757600021511*cos(theta1)*sin(theta2))/81129638414606681695789005144064 + cos(theta2)*sin(theta1)) + sin(theta3)((4967757600021511*cos(theta1)cos(theta2))/81129638414606681695789005144064 - sin(theta1)*sin(theta2))) + cos(theta4)(cos(theta1) - (4967757600021511*cos(theta3)((4967757600021511*cos(theta1)*cos(theta2))/81129638414606681695789005144064 - sin(theta1)*sin(theta2)))/81129638414606681695789005144064 + (4967757600021511*sin(theta3)((4967757600021511*cos(theta1)*sin(theta2))/81129638414606681695789005144064 + cos(theta2)*sin(theta1)))/81129638414606681695789005144064)

(4967757600021511*sin(theta2)sin(theta3))/81129638414606681695789005144064 - (4967757600021511*cos(theta2)*cos(theta3))/81129638414606681695789005144064 - cos(theta4)((4967757600021511*cos(theta2)cos(theta3))/81129638414606681695789005144064 - (4967757600021511*sin(theta2)*sin(theta3))/81129638414606681695789005144064 + 4967757600021511/81129638414606681695789005144064) + sin(theta4)(cos(theta2)*sin(theta3) + cos(theta3)*sin(theta2)) + 122597380068651197257713859414983140362437055831/533996758980227520598755426542388028650676130589163192486760401955554931445160137505740521734144

most of its elements are zero but i cannot get clean integers

[Walter Roberson]

syms theta1 theta2 theta3
E = (sym('24678615572571482867467662723121')*sin(theta1))/sym('6582018229284824168619876730229402019930943462534319453394436096') + (sym('4967757600021511')*cos(theta3)*(cos(theta1)*sin(theta2) + (sym('4967757600021511')*cos(theta2)*sin(theta1))/sym('81129638414606681695789005144064')))/sym('81129638414606681695789005144064')

E = 

vpa(E)

ans = 

Those are not zero.

Enew = mapSymType(expand(E), 'rational', @(x) piecewise(abs(x) < 1e-20, 0, x))

Enew = 

vpa(Enew)

ans = 

[Walter Roberson]

However, in any case like this, you should be working on the problem from much further back.

When you are building the expressions to be calculated on, enclose any non-integer floating point constant in a call to sym() . Even if it is just 0.5, use sym(0.5) . Do this even for integers if you are performing calculations on them, such as sqrt(sym(1)/2) or sqrt(sym(1)/sym(2)) . Also do it if you are working with larger integers.

Do this even with fractions. For example, sym(1)/sym(7) or sym(1)/7 . If you use sym(1/7) instead then MATLAB will often correctly guess which fraction you wanted, but it does not always get the value right.

If you have floating point constants of the form number1Enumber2 then replace that with sym(number1)*10^sym(number2) . For example, 1.23e6 -> sym(1.23)*10^(sym(6))

Any integer that exceeds 2^52 and cannot be expressed in the above sym(A)*10^sym(b) form with smaller integer, must be expressed with sym() and quoted values. For example, sym('24678615572571482867467662723121')

Enclose any reference to pi in a call to sym(), so for example sin(sym(pi)/3)

I usually find it easier to write a small function to help, such as

Q = @(x) sym(x);
Pi = Q(pi);
sin( Q(5) * Pi / Q(7))

If you take these steps, there will still be some cases where large numbers like the ones you saw will be generated, but the incidence will be greatly reduced.