Why do I get incorrect output when I run a command like "number*sum(x{1}(1:end).^integer)" in MATLAB 7.12 (R2011a)?

Question

MathWorks Support Team on 12 Apr 2012

I get incorrect output (ans = 0) when I run the following code:

   x{1} = [1 2 3 4 5]; 
mySum = sum( (x{1}(1:end)).^2 )
1 * sum( (x{1}(1:end)).^2 )

The result must match (ans = 55):

sum( (x{1}(1:end)).^2 )

The only difference is the 1 which is pre multiplied.

This appears to be an error in the parsing mechanism.

Answer 1

MathWorks Support Team on 12 Apr 2012

This bug has been fixed in Release 2012a (R2012a). For previous product releases, read below for possible workarounds:

There is a bug in the way MATLAB 7.10 (R2010a) parses the commands.

Possible workarounds are as follows:

1) Use the undocumented FEATURE function

feature accel off

2) Use POWER instead of the operator.

1 * sum(power( (x{1}(1:end)),2) )

instead of

1 * sum( (x{1}(1:end)).^2 )

NOTE: This bug seems to be present only if

1) The power is integer.

2) END is used instead of the length of the vector.

Why do I get incorrect output when I run a command like "number*su​m(x{1}(1:e​nd).^integ​er)" in MATLAB 7.12 (R2011a)?