EXAMPLES ON SELECTION INTENSITY, TEMPARATURE PREDICTION AND DAYLEGTH

Solutions may be more or less indicated.

Task 1
Order statistics is useful, it is ranked expectations in a sample from a standard normal distribution. Make a list of the individual expected values of the 10 top ranking in a sample of 1000.
Hint: Use program ?.exe...

 

Task 2

Hi, for prediction of selection intensities I need the following selection intensities, please find them out for me with an error less than <0.001!

a). Best 2000 from 10000

b). Best 0.0000571 from a very big number

c). Best 7 from 3658

d)  Best 2 from 11.

Lead: use SELEINT2!

Solution:

App.                     Burr.         Exact              error

  a)    1.39981             1.39967                                <0.001

  b)   G = i ; 4.09066,  truncation limit = 3.85825, variance = 0.04930

  c)    3.18296             3.16057         3.161102        <0.0001

  d)    exact (from table actually) 1.3241

Task 3
A truncation of the 1 percent most extreme tail in a normal distribution is made. What is the variance in the truncated tail?
Hint: Use program ?.exe...

Task 4
Compare truncation selection and optimal utilization from an infinite normal distribution for a Nr =0.1; 0.5 and 0.9 considering selection intensity (=gain) and variance.
Hint: Use program ?.exe...

Task 5
You spend the 24 of March 2001 at Mekijärvi and wonders about the temperature and light environment.
Mekrijärvi 62°46'N 30°59'E 150 m

Solution: TEMPPRED.EXE gives the predicted temp to -3.00 and average heat sum to 988
DAGL.EXE gives day to 12.69 and night 11.31 (with sun "below" the horizon).

There are more problems and examples in the Word7 document.
Last edited by Dag Lindgren 04-04-28, using a text file from before 1998.