M-function in a M-script

Question

3 个投票

When I write M-scripts (It's a script, not an M-Function), sometimes I want to have a quick and dirty function, such as:

test1.m

a=1;
b=MyFunction(a)
Function y=MyFunction(x)
y=x/2;
end

But it is not allowed because "FUNCTION keyword is invalid heare". I have to create another M-file called MyFunction.m to define MyFunction.

Why is that? What is the rational behind it?

I know I can work this way:

test2.m

function test2
a=1;
b=MyFunction(a)
end
function y=MyFunction(x)
y=x/2;
end

Although no need to specify the argument for function test2, still, I have to add the first line for function test2, add an "end" for function test2. And worst of all, no easy access to base workspace data.

0 个评论
显示 -2更早的评论隐藏 -2更早的评论

请先登录，再进行评论。

请先登录，再回答此问题。

Follow Question

Answer 1

Sean de Wolski 2011-5-18

0 个投票

That was covered in depth on CSSM.

http://www.mathworks.com/matlabcentral/newsreader/view_thread/298837

4 个评论
显示 2更早的评论隐藏 2更早的评论

Walter Roberson 2015-7-26

Let me think: Fortran... C... C++... VBA... java 7 and earlier... Nope, none of those have functions as first class objects that can be defined almost anywhere.

Scripting is like leaving broken glass on the floor. If you know where it is and always remember to walk around it, it isn't going to be a problem.

Saying that functions with dozens of arguments are "as bad or worse" is like saying that if you provide people with pieces of metal and grindstones then that is worse than glass on the floor because some people might leave dangerous metal burrs on the things they shape.

I accidentally broke a bottle on my back walk yesterday. I swept it up, but I couldn't get at the slivers in some of the cracks, and it's pretty likely that my feet will discover a missed sliver at some point.