Hello every body out there.
Kohei Yoshida wrote :
On Tue, 29 Mar 2005 00:18:11 -0500, Jordan Macdonald <[EMAIL PROTECTED]> wrote:
Just wondering, When do you think the Statistical data annalisis tools, especially the regression line for the spreadsheet chart, will be done?
[...]
With regard to the statistical data analysis tools, I don't think, please correct me if I'm wrong, anybody in the core team is looking into it, as it is one of the tasks they hope to be fulfilled by the community. I have in the past looked into it, but I just realized that my understanding of computational statistics is not good enough to tackle this issue at this point. But it is one of the things I want to go back to at some point in the future perhaps after I brush up my computatoinal statistics knowledge (right now I'm too busy studying numerical analysis and optimization for the solver component).
Of course, it someone else is interested in investigating into this *today*, nothing will stop him/her from doing it (in fact, I'm sure s/he will get any needed assistance along the way).
I have already work on some regression stuff, that you can find on issue 25478. The code already available on this issue can fit linear functions, exponential functions, logarithmic functions, potential functions and polynomial functions. It gives the computed functions, not only the curve. In fact, my code do not take care on the curve, because it is quite easy to draw a curve when you have got the functional to draw it.
The last code I have send on issue 25478 do not need many things before being an UNO component, in fact I have just to do some minor improvements before working on the interface but, well, I have not take the time do to this...
Please, take a look on the code I let on issue 25478 (the code I have send January 4, 2005) and tell me what you do think of it.
Best regards.
Yoann LE BARS,
alias Le Farfadet Spatial--------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
