Re: [ferret_users] Calculating Probability Density Function

[archana shrestha <archanamet@xxxxxxxxx>]

Dear Ansley,
Thank you very much for your suggestions. I have gone
through the statistic demo, but I think I have to
calculated the seasonal variance. Therefore I was
trying to get the seasonal time series, but I got
problem which I posted in my earlier email and
resending it again below.Can you help me in solving
this? 
........
Dear All,

I was trying to make time series of JJAS average (from
merged file). I found an example in the Mail Archive
for JJA. I followed the example and get the list of
JJAS time series. But the value for each year in this
time series is different from that I calculated for
the individual year (individual file). For example,
from the time series data JJAS average for 1961 was
14.17, while from induvidual year it was 12.83. Up to
the COMPRESS output, it matches with individual years
data. I would be grateful if anyone can solve this
problem. The method that I followed and the output are
as follows:

use ATM-ECHAMRF1961-89 !monthly data
let L_indx    = L[GT=rt] 
let yr_months = L_indx- 12*INT((L_indx-1)/12) 
list yr_months 
!masking for the months JJAS
let mask_other_months = IF yr_months GT 5 AND
yr_months LT 10 THEN 1 
!area of interest is x=80:92,y=26:30
let rt_masked = rt[d=1,x=80:92@ave,y=26:30@ave ]*
mask_other_months  
let rt_compr  = COMPRESSL(rt_masked) 
list rt_compr
define axis/t=1:116:1 t116  ! 4 months 29 yrs
=4*29=116
let rt_jjas_all = rt_compr[gt=t116@ASN]
list rt_jjas_all ! it gives correct values as follows
VARIABLE : RT_COMPR[GT=T116@ASN]
             FILENAME : ATM-ECHAMRF1961-89.des
             SUBSET   : 116 points (T)
 1     /   1:  12.24
 2     /   2:  17.95
 3     /   3:  10.47
 4     /   4:  10.72
 5     /   5:   7.39
 6     /   6:  16.71
 7     /   7:   9.40
 8     /   8:   7.22
 9     /   9:  17.95
.................
....
116

define axis/t=1:116:4 t29  
let rt_jjas_av  = rt_jjas_all[gt=t29@AVE] 
define axis/t=1961:1989:1/units=years tyear
let rt_jjas  = rt_jja_av[gt=tyear@ASN] 
list rt_jjas ! these values do not match with
individual year calculation 

VARIABLE : RT_JJAS_AV[GT=TYEAR@ASN]
             FILENAME : ATM-ECHAMRF1961-89.des
             SUBSET   : 29 points (T (YEARS))
 1961   /  1:  14.17
 1962   /  2:  11.19
 1963   /  3:  12.48
 1964   /  4:  13.24
...................
..............
 1989..............

Individual year values are

1961  12.83
1962  10.18
....etc


--- Ansley Manke <ansley.b.manke@xxxxxxxx> wrote:

> Archana, 
> You might look at this, making a seasonal mask
>
http://ferret.pmel.noaa.gov/FERRET_17sep07/FAQ/analysis/seasonal_masking.html
> 
> You would need to change the definitions in the
> examples to get the 
> months you want to choose.
> 
> Also run this script, which shows how to define
> probability density 
> functions.
> 
> yes? go statistics_demo
> 
> It creates data on an x axis, and shows how to
> define statistics on that 
> variable ARG.
> 
> 
> Ansley
> 
> 
> archana shrestha wrote:
> > Dear Ferreters,
> > I have ATM-ECHAMRF1961-89.des produced by merging
> 29
> > nc files using descripter.  I want to make
> probability
> > Density function (PDF) for seasonal rainfall for
> DJF,
> > JJAS, ON and MAM. My variable description is as
> below:
> >
> > name     title            I      J         K      
>   L
> > RT       Total Precip   1:118   1:109     ...     
> 
> > 1:348
> >
> > I have looked at the Mail archive, but could not
> get
> > the appropriate answer. Do I need to first
> calculate
> > standard deviation and mean? If so how can I
> calculate
> > seasonal variance? Do I need to average my
> variable
> > over certain area? Can anyone help me to do this
> > analysis?
> >
> > Thank you in advance
> > Archana
> >
> >
> >      
>
____________________________________________________________________________________
> > Be a better friend, newshound, and 
> > know-it-all with Yahoo! Mobile.  Try it now. 
>
http://mobile.yahoo.com/;_ylt=Ahu06i62sR8HDtDypao8Wcj9tAcJ
> >   
> 


Ms. Archana Shrestha
  Meteorologist
  Department of Hydrology and Meteorology
  Babarmahal
  P.O. Box: 406, Kathmandu, Nepal
  Tel: 977-1-4255920
  Fax:977-1-4254890
  web:www.dhm.gov.np



      ____________________________________________________________________________________
Be a better friend, newshound, and 
know-it-all with Yahoo! Mobile.  Try it now.  http://mobile.yahoo.com/;_ylt=Ahu06i62sR8HDtDypao8Wcj9tAcJ