[ncl-talk] regCoef(x,y)

[Xi Chang]

Hi Dennis,

So, is it correct if I plot like this:

prob = (1.-*betainc*(df/(df+tval^2),df/2.0,b))*100.

In the example that you pointed out, it is not really clear for me.


Chang


On Thu, Mar 17, 2016 at 9:32 PM, Dennis Shea <shea at ucar.edu> wrote:

> See example 2  of
>
> http://www.ncl.ucar.edu/Document/Functions/Built-in/betainc.shtml
>
> ---
> If you are testing at the 5% level (0.05) and prob is < 0.05, it is
> significant.
>
>
>
> On Thu, Mar 17, 2016 at 2:10 PM, Xi Chang <xi.chang01 at gmail.com> wrote:
>
>> Hallo NCL,
>>
>> I looked at exampled 2
>> <https://www.ncl.ucar.edu/Document/Functions/Built-in/regCoef-1.shtml>,
>> which is exactly my case. But it's not clear after getting "prob" defined
>> as :
>>
>> *prob= betainc*(df/(df+tval^2),df/2.0,b)
>>
>> How could I determine that only signals that are statistically
>> significant at 95% will be shown in the plot.
>>
>> any help?
>> Thanks
>>
>> Chang
>>
>>
>>
>> _______________________________________________
>> ncl-talk mailing list
>> ncl-talk at ucar.edu
>> List instructions, subscriber options, unsubscribe:
>> http://mailman.ucar.edu/mailman/listinfo/ncl-talk
>>
>>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.ucar.edu/pipermail/ncl-talk/attachments/20160801/83f2f270/attachment.html