Re: Am I right with coherence probility

Got it. Thanks.
I have read them, but still confused. So ...

2014-02-26



Xueming Zhu 朱学明
Key Laboratory of Research on Marine Hazards Forecasting (LoMF), SOA
National Marine Environmental Forecasting Center (NMEFC)
No.8, Dahuisi Road, Haidian District, Beijing, 100081
People's Republic of China
Tel:+86-10-82481923



发件人：Dennis Shea <shea@ucar.edu>
发送时间：2014-02-26 06:06
主题：Re: [ncl-talk] Am I right with coherence probility
收件人："朱学明"<557xiaozhu@163.com>,"ncl-talk@ucar.edu USERS"<ncl-talk@ucar.edu>
抄送：

Well, you should read the two references.

I am not a statistician but I think you should use
the returned coherence and not sqrt(...)

Good luck


On 2/25/14, 2:46 PM, 朱学明 wrote:
> Hi Dennis,
> Thanks for your reply. Can you make sure for me which one is correct? Should I use sqrt or not?
>
> Am I rigth with the line
>> phase=where(sqrt(spec@coher).le.spec@coher_probability(1),phase@_FillValue,phase),
>> or
>> phase=where(spec@coher.le.spec@coher_probability(1),phase@_FillValue,phase)
>> in order to get the coherence satisfy the 95% confidence level?
>
>
> 2014-02-26
>
>
>
> Xueming Zhu 朱学明
> Key Laboratory of Research on Marine Hazards Forecasting (LoMF), SOA
> National Marine Environmental Forecasting Center (NMEFC)
> No.8, Dahuisi Road, Haidian District, Beijing, 100081
> People's Republic of China
> Tel:+86-10-82481923
>
>
>
> 发件人：Dennis Shea <shea@ucar.edu>
> 发送时间：2014-02-26 05:40
> 主题：Re: [ncl-talk] Am I right with coherence probility
> 收件人："朱学明"<557xiaozhu@163.com>,"ncl-talk@ucar.edu USERS"<ncl-talk@ucar.edu>
> 抄送：
>
> I would say "yes"
>
> On 2/24/14, 2:52 PM, 朱学明 wrote:
>> Sorry for the corrections.
>>
>>
>> Hello all,
>> I have two time series data and want to do cross spectrum analysis. Can you help me to check whether I have got the right coherence probility? My codes are as follow,
>>
>> fn = sec1+"_daily.dat.bak"
>> data = asciiread(fn,(/nline,2/),"float")
>> xn = data(:,1)
>> fn = sec2+"_daily.dat.bak"
>> yn = data(:,1)
>> wgtt = filwgts_lanczos (301,2,1/400,1/80,1.0)
>> XN = wgt_runave (xn,wgtt,0)
>> YN = wgt_runave (yn,wgtt,0)
>> delete([/xn,yn/])
>> xn = XN((nwgt-1)/2:(dimsizes(XN)-(nwgt-1)/2-1))
>> yn = YN((nwgt-1)/2:(dimsizes(YN)-(nwgt-1)/2-1))
>> spec = specxy_anal(xn,yn,0,5,0.1)
>> dataL=new((/2,dimsizes(spec@coher)/),typeof(spec@coher))
>> dataL(0,:)=sqrt(spec@coher)
>> dataL(1,:)=spec@coher_probability(1)
>> phase=spec@phase
>> phase@_FillValue= 1e20
>> phase=where(sqrt(spec@coher).le.spec@coher_probability(1),phase@_FillValue,phase)
>>
>> plot=gsn_csm_xy2(wks,spec@frq,dataL,phase,resL,resR); create plot
>>
>>
>>
>> Am I rigth with the line
>> phase=where(sqrt(spec@coher).le.spec@coher_probability(1),phase@_FillValue,phase),
>> or
>> phase=where(spec@coher.le.spec@coher_probability(1),phase@_FillValue,phase)
>> in order to get the coherence satisfy the 95% confidence level?
>>
>> Thanks.
>>
>> 2014-02-25
>>
>>
>>
>> Xueming Zhu 朱学明
>> Key Laboratory of Research on Marine Hazards Forecasting (LoMF), SOA
>> National Marine Environmental Forecasting Center (NMEFC)
>> No.8, Dahuisi Road, Haidian District, Beijing, 100081
>> People's Republic of China
>> Tel:+86-10-82481923
>>
>>
>>
>> _______________________________________________
>> ncl-talk mailing list
>> List instructions, subscriber options, unsubscribe:
>> http://mailman.ucar.edu/mailman/listinfo/ncl-talk
>>
>>
>> _______________________________________________
>> ncl-talk mailing list
>> List instructions, subscriber options, unsubscribe:
>> http://mailman.ucar.edu/mailman/listinfo/ncl-talk

_______________________________________________
ncl-talk mailing list
List instructions, subscriber options, unsubscribe:
http://mailman.ucar.edu/mailman/listinfo/ncl-talk
Received on Tue Feb 25 15:08:39 2014