Re: a small fix for hist.m

[Doug Stewart]

On Wed, Dec 1, 2010 at 8:30 AM, Ben Abbott <address@hidden> wrote:
> On Dec 1, 2010, at 8:24 AM, Ben Abbott wrote:
>
>> On Nov 30, 2010, at 10:39 PM, Doug Stewart wrote:
>>
>>> On Tue, Nov 30, 2010 at 9:33 PM, Doug Stewart <address@hidden> wrote:
>>>> On Tue, Nov 30, 2010 at 9:19 PM, Ben Abbott <address@hidden> wrote:
>>>>> On Nov 30, 2010, at 7:35 PM, Doug Stewart wrote:
>>>>>
>>>>>> I am now retired from teaching, and am going to spend some time
>>>>>> improving the help section of some M files.
>>>>>>
>>>>>> I am learning as I go so please go easy on me :-)
>>>>>>
>>>>>>
>>>>>> I have got hg working.
>>>>>> I have got the tip and made a change to hist.m
>>>>>> I have attached the diff file.
>>>>>>
>>>>>> I am also suppose to create one more document about the changes.
>>>>>> Can someone show me how to do this last part
>>>>>>
>>>>>> Doug Stewart
>>>>>>
>>>>>> PS. This is just a small fix for me to learn on.
>>>>>> <hist.diff>
>>>>>
>>>>>
>>>>> I see one minor format mistake. The output and equals sign should be 
>>>>> enclosed in the curly brackets.
>>>>>
>>>>> Forgive the micro-management, but I assume you'll be adding the various 
>>>>> modes of operation as well. Perhaps something like ...
>>>>>
>>>>> ## @deftypefn {Function File} {} hist (@var{y})
>>>>> ## @deftypefnx {Function File} {} hist (@var{y}, @var{x})
>>>>> ## @deftypefnx {Function File} {} hist (@var{y}, @var{nbins})
>>>>> ## @deftypefnx {Function File} {} hist (@var{y}, @var{x}, @var{norm})
>>>>> ## @deftypefnx {Function File} {[nn, xx] =} hist (@dots{})
>>>>>
>>>>> Ben
>>>>
>>>> Thank you Ben for the help (mico-managment :-) )
>>>> Yes   after I get the process figured out I hope to do much more.
>>>> And yes I do know about passing the colors through to bar()
>>>>
>>>> I will resubmit this one.
>>>> Doug
>>>
>>> here is the new diff
>>> Is there anything else I should do?
>>> <hist.diff>
>>
>> The help text can be improved by changes such as ....
>>
>> -## With one vector input argument, plot a histogram of the values
>> +## With one vector input argument, @var{y}, plot a histogram of the values
>>
>> If you're up to a bug fix, it looks to me like the example below does not 
>> produce the correct result.
>>
>> y = randn (1, 1000);
>> x = [0.1, 0.2, 0.5, 0.7 0.9];
>> hist (y, x)
>>
>> Matlab gives the attached result. Its not immediately clear to me how the 
>> bin widths are determined.
>>
>> Ben
>> <hist.pdf>
>
> I need my morning coffee ... but it looks like the edges of the bins are 
> centered between adjacent values for "x", with the outer most edges extending 
> to encompass all values of "y".
>
> Ben
>
>

Yes that is what I see now.

So for non uniform bin spacing
  - find max and min
  - find point half way between each given point
  - construct a group of ranges from the above
  - count how many are in each of these non uniform ranges.
Does this look like something we should implement?
 If yes I will work on it latter.
Doug