Overwriting Solve Block tolerances, TOL and CTOL.

[KAL_1625]

I am wondering , what is a precise tolerance and what is a loose tolerance. I would imagine that from 10^-1 to 10^-9, loosest tolerance would be 10^-1 and tightest would be 10^-9. Please correct me if I am wrong. 
I also had question if I could overwrite tolerance even further? I want to loosen my tolerance even further since my solve block is not converging for 10^-1 (Assuming it being the loosest tolerance of all). Thank you in advance.

[Werner_E · Accepted solution]

Yes, the larger the value of CTOL or TOL, the "looser" the tolerance.

You can always set the system variables TOL and CTOL in the worksheet like you assign normal variables. So you always can write TOL:=0.5 in fromt of a calculation where you want a very loose tolerance.

But I guess that if a solve block fails with a tolerance of 10^-1, it would not help to set TOL or CTOL even larger than that. I would rather double check the equations in use and the initial guess values.
And even if the solve block gives you a solution with a tolerance as high as 0.5 or even 1, its doubtful if this "solution" can be of any relevance.

If you use a solve block with "find" you may try to use "minerr" instead of "find", too.

[KAL_1625]

Thanks for your response. it is very clear for me now. My system of equations has 17 equation and 17 unknowns, so it takes a lot of time to solve those equation. Sometimes I get the solution and sometimes I don't for same set of equation. I takes 20 to 30 mins of solving and shows me the error to change the guess or TOL values. In this case I don't need my solution to be at a tolerance of 10^-1. I can go loose with it. Like you said I tried with minerr, but sometime I got highly unrealistic solutions with those. 
I guess overwriting TOL should be helpful for me. Thanks you very much.