Sum through an array until a value is reached, then continue

Assume you have 2 arrays x = linspace(0.001,0.005,1000); y = linspace(0,1,1000);

I want to start at the beginning of x, and sum consecutive columns until the sum = 0.1, then take the average of y across that range.... then continue doing this until the end of x

[madhan ravi]

give a short example of your desired result

[Anthony]

my desired result is two new arrays that have the averages of y across the given intervals of x.

Edit: Maybe easier to think of x as a length measurement - as in, I want the average value of y for every 0.1m interval

[Bruno Luong]

Description inaccurate and contradictory.

It looks like the "cumulative sum" is actually y (100 elements).

x never reaches 0.1.

Cumulative sum on x and y never give 0.1 crossing at 100 elements.

[Anthony]

It doesn't have to be 100 elements, just using an example. In my situation, I do not know how many cells of x I need to sum over to reach 0.1.

I should also say that it will not likely be exactly 0.1 ... so having a margin of error, or rounding numbers would be useful.

[Bruno Luong]

"With the values given, the first 100 columns would sum to 0.1"

No body know where the 100 comes from. Anyone is guessing and people just wastes theirs times for you.

Not give example is better than give one that doesn't make sense.

[Anthony]

Thanks for trying Bruno, you're a peach!

Answer 1

[Matt J]

   c=cumsum(x);

   G=discretize(c, 0:0.1:c(end));

   result = splitapply(@mean, y, G),

[Anthony]

This looks like it working how I want... Thanks a lot!

Answer 2

[madhan ravi]

mean(y(cumsum(x)<=0.1))

[Anthony]

but this only gives one answer. From my previous edit: think of x as a length measurement - as in, I want the average value of y for every 0.1m interval

thus, what I want to get is average y between 0-10cm, then average y over the next 10cm interval, etc... then end result would be more than one number

[madhan ravi]

see the edited answer

[Anthony]

Thanks for trying to help, but this deosn't produce a correct output

[madhan ravi]

did you read Brunos comment?

[Guillaume]

In case you haven't noticed you were given another answer that matches your description.

"this deosn't produce a correct output" is totally useless if you don't tell us what the correct output is. Vagues descriptions that don't actually match the example you've given doesn't cut it.

As has been pointed out, your statement "the first 100 columns would sum to 0.1" is incorrect. cumsum(x) reaches 0.1 at column 86.

[Anthony]

Yes, I was incorrect about the 100 columns, as I stated. Regardless, I was quite clear, and someone else figured it out. Thanks for trying to help and taking the time for snide comments ;)

[Bruno Luong]

Yes 1 over 4 people figures it out ...by chance.

[Anthony]

Thanks again Bruno, you're the best!

Answer 3

[Guillaume]

sumx = cumsum(x);
meanybelowthreshold = mean(y(sumx <= 0.1));
meanyabovethreshold = mean(y(sumx > 0.1));