IF statement with multiple logical OR

조회 수: 3 (최근 30일)
George Holt
George Holt 2015년 12월 25일
댓글: George Holt 2015년 12월 25일
I'm running a function that solves a given PDE using Euler's method
I attempt to plot the numerical solutions to the PDE at specific times (0.1, 0.5, 1.0 and 2.0) by testing equalities for t in an IF statement
However, only plots for t == 0.1 and t == 0.5 are plotted for some reason
I've tried debugging with breakpoints and the IF statement tests true when t == 0.1 and 0.5, but false when t == 1 and 2, even though the variable t in the workspace == 1 and 2
The relevant code is:
t = 0;
dt = 0.1;
for a = 1 : timesteps
for b = 1 : nodes
u = %solve PDE;
end
t = t + dt;
round(t,1);
if t == 0.1 || t ==0.5 || t == 1.0 || t == 2.0
figure
plot(t,u)
end
end
  댓글 수: 2
John D'Errico
John D'Errico 2015년 12월 25일
NEVER test for exact equality of a floating point number. That test will fail most of the time. Note that 0.1 is not representable exactly in a binary arithmetic.
George Holt
George Holt 2015년 12월 25일
Many thanks! I tested for inequalities instead and that works

댓글을 달려면 로그인하십시오.

채택된 답변

Jan
Jan 2015년 12월 25일
This line does not perform ynathing, because theoutput is not stored:
round(t,1);
You cannot expect, that 0.1+0.1+0.1 is exactly 0.3, because the most decimal floating point values do not have an exact representation in the binary system, such that rounding errors occur.
Better use an arbitrary, but matching threshold:
if any(abs(t - [0.1, 0.5, 1.0. 2.0]) < 1e8)
figure
plot(t,u)
end
Add a meaningful comment to explain the size of the threshold.
Since R2015a:
if ismembertol(t, [0.1, 0.5, 1.0. 2.0])
  댓글 수: 1
George Holt
George Holt 2015년 12월 25일
I used
if (0 < t) && (t < 0.2) || (0.4 < t) && (t < 0.6) %etc
but your solution is much nicer!
Thanks

댓글을 달려면 로그인하십시오.

추가 답변 (0개)

카테고리

Help CenterFile Exchange에서 Boundary Conditions에 대해 자세히 알아보기

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by