Re: BUG #13619: regression functions return Null

Oh, and don't reply to people individually. I'm going to move this to
-general with one last copy for -bugs even though it is not one.

On Mon, Sep 14, 2015 at 3:01 PM, David G. Johnston <
david(dot)g(dot)johnston(at)gmail(dot)com> wrote:

> On Mon, Sep 14, 2015 at 2:55 PM, Biswadeep Banerjee <
> biswadeep(dot)banerjee(at)actifio(dot)com> wrote:
>
>> Hi David,
>> May be I gave a wrong example to explain my issue.
>>
>> Below is a sample query that I have been trying to get the slope and
>> intercept. The inner query basically gives me the time vs capacity, ie. x
>> and y data showing the actuals. Based on the x, y value, I calculate slope
>> and intercept value, as in the outer query.
>>
>> But it looks like am back to your comment as it mentions I am asking for
>> slope and intercept of a single point. Could you provide me some examples
>> that I can use as a reference.
>> As a reference, I am following similar example as in
>> http://stackoverflow.com/questions/20490756/linear-regression-with-postgres
>>
>> select x, y, regr_slope(y,x), regr_intercept(y,x)
>> from (
>> ​/* doesn't matter */
>>
>> ) i
>> group by x, y
>> order by x asc, y asc
>>
>
> ​You are grouping on (x, y)​
>
> ​ and then passing this single POINT into a function that requires
> multiple points in order to calculate the slope and intercept of a LINE
> (i.e., something requiring two points to describe) and are confused why it
> is giving you NULL...
>
> ​I have answered your question - you are using the functions incorrectly.
>
> This is operator error, not a bug.
>
> You have already found a reasonably good example of how these functions
> can be used. Given I have never used them myself I do not have anything
> better to offer.
>
> David J.
> ​
> ​
>