## the problem of using 'case'

### dcydhb dcydhb (view profile)

on 13 May 2019
Latest activity Commented on by dcydhb dcydhb

on 13 May 2019

### Stephen Cobeldick (view profile)

i want to use a loop of omega and when the omega has diffreent value ,the m0,m1,m2...have the different value,so in the loop i use the 'case' as this
for omega=0.1:0.1:2;
switch omega
case 0.100000;
m0=0.031949;
m1=3.141268;
m2=6.283023;
m3=9.424670;
m4=12.566289;
m5=15.707898;
m6=18.849502;
m7=21.991102;
m8=25.132701;
m9=28.274298;
m10=31.415894;
case 0.200000;
m0=0.063931;
m1=3.140293;
m2=6.282536;
m3=9.424345;
m4=12.566046;
m5=15.707703;
m6=18.849339;
m7=21.990963;
m8=25.132579;
m9=28.274190;
m10=31.415797;
and there are still some codes,but in the result,when omega=0.2,it still use the m0,m1,m2...of case 0.100000;
so when i change it as
case abs(omega-0.200000)<0.01;
it still has the problem,and why,thanks a lot!!!

#### 1 Comment

Stephen Cobeldick

### Stephen Cobeldick (view profile)

on 13 May 2019
Numbering variable names is a sign that you are doing something wrong. You should simply store all of those scalar numerics in one numeric array.

R2014a

### Stephen Cobeldick (view profile)

on 13 May 2019
Edited by Stephen Cobeldick

### Stephen Cobeldick (view profile)

on 13 May 2019

for omega = 0.1:0.1:2;
switch omega
case 0.1
disp('one')
case 0.2
disp('two')
end
end
displays this in the command window:
one
two
Because of floating pont error in those numbers, this would be an extremely unreliable way to write code. Do NOT do this!
Your second method can be done quite easily, by simply thinking about the actual case value that you actually want switch to compare (hint: your case conditions are equal to true or false, but omega is not equal to true or false):
for omega = 0.1:0.1:2;
switch true % this is the value for a valid CASE.
case abs(omega-0.1)<0.01;
disp('one')
case abs(omega-0.2)<0.01;
disp('two')
end
end
displays this in the command window:
one
two

dcydhb dcydhb

### dcydhb dcydhb (view profile)

on 13 May 2019
i have accepted as it is wonderful and i was curious that
switch true % this is the value for a valid CASE.
case abs(omega-0.1)<0.01;
disp('one')
you must use the
switch true % this is the value for a valid CASE.
rather than
switch omega
why?
Walter Roberson

### Walter Roberson (view profile)

on 13 May 2019
When you use
switch numeric_value
case another_numeric_value
then the test is
if numeric_value == another_numeric_value
So let us follow that through with your proposed version with switch omega:
if omega == (abs(omega-0.1) < 0.01)
The right hand side is a logical value, false or true, convertable to numeric 0 or 1, and you want to compare that to omega. That can only possibly succeed if omega is 0 or 1 . In the case where omega is 1 then abs(1-0.1) is 0.9 and < 0.01 of that is false. In the case where omega is 0, then abs(0-0.1) is 0.1 and < 0.01 of that is false. Therefore switch omega cannot work.
if true == (abs(omega-0.1) < 0.01)
which can be true if omega is in the range approximately 0.1-0.01 to 0.1+0.01
dcydhb dcydhb

### dcydhb dcydhb (view profile)

on 13 May 2019
great explanation and thank you all a lot!!!