[ncl-talk] question on dim_avg_n and dim_stddev_n

Thu Nov 9 15:05:02 MST 2017

I fixed it Dennis. Thanks. It was because I had unequal samples in the two
experiments being compared.

On Thu, Nov 9, 2017 at 3:01 PM, Dennis Shea <shea at ucar.edu> wrote:

> Hi Andrew ...
>
> Was this answered?
> Did you find the problem?
>
> "Could it be there are missing values for certain storms and cycles?"
> .... quite possibly.
>
> D
>
> On Tue, Oct 31, 2017 at 4:18 PM, Andrew Kren - NOAA Affiliate <
> andrew.kren at noaa.gov> wrote:
>
>> Dear ncl-talk,
>>
>> I am computing a paired t test between control experiments sensitivity
>> experiments to look at hurricane track, wind, and intensity errors. I
>> compute the standard deviation of the differences and the avg of the
>> differences, as well as a 95% confidence interval based on the differences.
>>
>> Attached is my code, see lines starting on 1127. Is this the correct way
>> to compute the standard deviation and average of the differences between
>> experiments? My data is a function of (storm case, forecast cycle, forecast
>> hour). Reason I ask is that when I do this procedure, the average
>> differences I get using this method as a function of forecast hour (which
>> is what I want) is this:
>>
>> Variable: track_error_diff_avg (subsection)
>>
>> Type: float
>>
>> Total Size: 116 bytes
>>
>>             29 values
>>
>> Number of Dimensions: 1
>>
>> Dimensions and sizes: [29]
>>
>> Coordinates:
>>
>> Number Of Attributes: 1
>>
>>   _FillValue : 9.96921e+36
>>
>> (0) 3.574069
>>
>> (1) 5.169381
>>
>> (2) 4.654985
>>
>> (3) 1.50052
>>
>> (4) 3.559848
>>
>> (5) 3.355001
>>
>> (6) 4.764093
>>
>> (7) 8.347384
>>
>> (8) 7.4192
>>
>> (9) 8.69336
>>
>> (10) 12.72247
>>
>> (11) 17.61452
>>
>> (12) 17.19171
>>
>> (13) 11.51769
>>
>> (14) 9.379377
>>
>> (15) 11.44438
>>
>> (16) 7.419726
>>
>> (17) -3.023807
>>
>> (18) -18.86928
>>
>> (19) -3.98752
>>
>> (20) -4.023374
>>
>> (21) -14.1253
>>
>> (22) -7.246421
>>
>> (23) -10.78448
>>
>> (24) -23.74559
>>
>> (25) -40.31917
>>
>> (26) -51.49041
>>
>> (27) -34.30735
>>
>> (28) -9.808663
>>
>>
>> However, when I compute the raw difference from the average track fields,
>> i.e. line 1154, I get these values:
>>
>>
>> Variable: track_error_diff_avg (subsection)
>>
>> Type: float
>>
>> Total Size: 116 bytes
>>
>>             29 values
>>
>> Number of Dimensions: 1
>>
>> Dimensions and sizes: [29]
>>
>> Coordinates:
>>
>> Number Of Attributes: 1
>>
>>   _FillValue : 9.96921e+36
>>
>> (0) 3.574068
>>
>> (1) 5.16938
>>
>> (2) 4.654984
>>
>> (3) 1.500519
>>
>> (4) 3.559845
>>
>> (5) 3.355003
>>
>> (6) 4.764091
>>
>> (7) 8.654808
>>
>> (8) 8.30986
>>
>> (9) 6.454712
>>
>> (10) 10.45447
>>
>> (11) 17.66672
>>
>> (12) 20.24161
>>
>> (13) 14.71471
>>
>> (14) 13.84648
>>
>> (15) 17.98428
>>
>> (16) 9.601624
>>
>> (17) -0.5063171
>>
>> (18) -14.034
>>
>> (19) -3.797974
>>
>> (20) -0.2883911
>>
>> (21) -11.9198
>>
>> (22) -9.015442
>>
>> (23) -6.522919
>>
>> (24) -61.46265
>>
>> (25) -74.77875
>>
>> (26) -40.00131
>>
>> (27) 52.12799
>>
>> (28) 76.7471
>>
>>
>> where the values start to deviate after index 8. Is there something I am
>> doing wrong? I can't seem to find out why if so. Could it be there are
>> missing values for certain storms and cycles? Any help is much appreciated.
>>
>> --
>> Andrew Kren
>> Research Scientist I, Global Observing Systems Analysis (GOSA) Group
>> NOAA ESRL Global Systems Division (Rm 3C515)
>> 325 Broadway, Boulder, CO 80305
>> <https://maps.google.com/?q=325+Broadway,+Boulder,+CO+80305&entry=gmail&source=g>
>> (303) 497-5418
>>
>> _______________________________________________
>> ncl-talk mailing list
>> ncl-talk at ucar.edu
>> List instructions, subscriber options, unsubscribe:
>> http://mailman.ucar.edu/mailman/listinfo/ncl-talk
>>
>>
>

-- 
Andrew Kren
Research Scientist I, Global Observing Systems Analysis (GOSA) Group
NOAA ESRL Global Systems Division (Rm 3C515)
325 Broadway, Boulder, CO 80305
(303) 497-5418
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.ucar.edu/pipermail/ncl-talk/attachments/20171109/17b93b80/attachment.html>