[ncl-talk] stat_dispersion and standard deviation

jbuzan jbuzan at purdue.edu
Mon May 11 17:11:20 MDT 2015


Thanks Dennis!

-Jonathan


> On Apr 24, 2015, at 4:50 PM, Dennis Shea <shea at ucar.edu> wrote:
> 
> ====
> 6.3.0
> 
> http://www.ncl.ucar.edu/current_release.shtml <http://www.ncl.ucar.edu/current_release.shtml>
> Under 'Bugs Fixed'
> 
> 
> stat_dispersion <http://www.ncl.ucar.edu/Document/Functions/Contributed/stat_dispersion.shtml> - The second returned element returned was the sample variance and not the sample standard deviation as is documented. In 6.3.0, the sample standard deviation will be returned. 
> 
> =====
> 
> %> ncl -V
> 6.3.0
> 
>   a = (/-20.46505, -4.795454, -1.454606,-2.677286, -1.868608, -0.4972475, 1.18715, -1.657706, -3.428875, -2.766053, -1.689298, -0.3371645, -0.640966, -0.7326741/)
> 
>   opt = True
>   opt at PrintStat = True
>   stata = stat_dispersion(a, opt )
> 
> 
> Copyright (C) 1995-2015 - All Rights Reserved
>  University Corporation for Atmospheric Research
>  NCAR Command Language Version 6.3.0          <====================
>  The use of this software is governed by a License Agreement.
>  See http://www.ncl.ucar.edu/ <http://www.ncl.ucar.edu/> for more details.
> (0)     
> (0)       ===> Robust Dispersion Statistics <===
> (0)     [0]            Mean=-2.98742
> (0)     [1]          StdDev=5.24311    <=====
> (0)     [2]             Min=-20.465
> (0)     [3]          LowDec=-20.465
> [SNIP]
> 
> 
> 
> On Fri, Apr 24, 2015 at 2:25 PM, jbuzan <jbuzan at purdue.edu <mailto:jbuzan at purdue.edu>> wrote:
> I don’t know if this has been fixed in 6.3.0, but there was a change in stat_dispersion and the standard deviation is no longer calculated.  it is the variance.
> 
> you can change:
> $NCARG_ROOT/lib/ncarg/nclscripts/csm/contributed.ncl
> (I renamed the file)
> 
> and commented out:
>      ; statx(1) = moment(1)             ; std. deviation
> 
> and replaced with:
>       statx(1) = stddev(x)             ; std. deviation
> 
> 
> -Jonathan
> 
> 
> 
>> On Apr 24, 2015, at 4:20 PM, Rainer Luptowitz <rlupt001 at ucr.edu <mailto:rlupt001 at ucr.edu>> wrote:
>> 
>> Hi ncl-talk,
>> 
>> Is the standard deviation computed in the "stats_dispersion" function actually a variance?
>> 
>> I test this and it appears to be variance instead of stats_dispersion as seen in with sample "a" below.
>> 
>> 
>> Variable: a 
>> Type: float
>> Total Size: 56 bytes
>>             14 values
>> Number of Dimensions: 1
>> Dimensions and sizes:	[ncl_join | 14]    ; I had joined 8 files trends with "addfiles" function
>> Coordinates:
>> Number Of Attributes: 0
>> (0)	-20.46505
>> (1)	-4.795454
>> (2)	-1.454606
>> (3)	-2.677286
>> (4)	-1.868608
>> (5)	-0.4972475
>> (6)	1.18715
>> (7)	-1.657706
>> (8)	-3.428875
>> (9)	-2.766053
>> (10)	-1.689298
>> (11)	-0.3371645
>> (12)	-0.640966
>> (13)	-0.7326741
>> 
>> ; ncl code compare variance to standard deviation
>> 
>> ; the above data was placed into an array named "a"
>> a = (/-20.46505, -4.795454, -1.454606,-2.677286, -1.868608, -0.4972475, 1.18715, -1.657706, -3.428875, -2.766053, -1.689298, -0.3371645, -0.640966, -0.7326741/)
>> 
>> 
>> ; Input
>> Var_a = variance(a)  ; calc variance
>> print(Var_a)
>> StdDev_a = stddev(a) ; calc standard deviation
>> print(StdDev_a)
>> opt = True
>> opt at PrintStat = True
>> Stats_a = stat_dispersion(a,opt)
>> 
>> 
>> ; Output
>> ; from individually calculating Variance
>> Variable: Var_a
>> Type: float
>> Total Size: 4 bytes
>>             1 values
>> Number of Dimensions: 1
>> Dimensions and sizes:	[1]
>> Coordinates:
>> (0)	27.49021
>> 
>> ; from individually calculating Standard Deviation
>> Variable: StdDev_a
>> Type: float
>> Total Size: 4 bytes
>>             1 values
>> Number of Dimensions: 1
>> Dimensions and sizes:	[1]
>> Coordinates:
>> (0)	5.243111
>> 
>> ; from Stats_Dispersion function.
>> (0)
>> (0)	   ===> Robust Dispersion Statistics <===
>> (0)	 [0]            Mean=-2.98742
>> (0)	 [1]          StdDev=27.4902    ; this appears to be variance since
>> (0)	 [2]             Min=-20.4651
>> (0)	 [3]          LowDec=-20.4651
>> (0)	 [4]          LowOct=-20.4651
>> (0)	 [5]          LowSex=-4.79545
>> (0)	 [6]     LowQuartile=-3.42888
>> (0)	 [7]          LowTri=-2.76605
>> (0)	 [8]          Median=-1.6735
>> (0)	 [9]         HighTri=-1.45461
>> (0)	 [10]   HighQuartile=-0.732674
>> (0)	 [11]        HighSex=-0.640966
>> (0)	 [12]        HighOct=-0.497248
>> (0)	 [13]        HighDec=-0.497248
>> (0)	 [14]            Max=1.18715
>> (0)	 [15]          Range=21.6522
>> (0)	 [16]     Dispersion=0.787633
>> (0)	 [17]    RMS Anomaly=5.05239
>> (0)	 [18]      #   Total=14
>> (0)	 [19]      #    Used=14
>> (0)	 [20]      # Missing=0
>> (0)	 [21]      % Missing=0
>> (0)	 [22]     Lower 0.1%=9.96921e+36
>> (0)	 [23]     Lower 1.0%=9.96921e+36
>> (0)	 [24]     Lower 5.0%=9.96921e+36
>> (0)	 [25]     Upper 5.0%=9.96921e+36
>> (0)	 [26]     Upper 1.0%=9.96921e+36
>> (0)	 [27]     Upper 0.1%=9.96921e+36
>> (0)	 [28]       Skewness=-2.57898
>> (0)	 [29]       Kurtosis=5.86411
>> 
>> 
>> What do you think?
>> 
>> Thank you for time and assistance,
>> 
>> Rainer Luptowitz
>> Grad Student 
>> Dept Earth Sciences
>> UC Riverside
>> rlupt001 at ucr.edu <mailto:rlupt001 at ucr.edu>
>> 
>> _______________________________________________
>> ncl-talk mailing list
>> List instructions, subscriber options, unsubscribe:
>> http://mailman.ucar.edu/mailman/listinfo/ncl-talk <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 <http://mailman.ucar.edu/mailman/listinfo/ncl-talk>
> 
> 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.ucar.edu/pipermail/ncl-talk/attachments/20150511/1e9fe34f/attachment.html 


More information about the ncl-talk mailing list