EXPLORATORY DATA ANALYSIS FROM DATA SET OF CORN SEED MATURATION

CARLOS ALBERTO SCAPIM, TEREZINHA APARECIDA GUEDES, JOSÉ WALTER PEDROZA CARNEIRO, ALESSANDRO DE LUCCA E BRACCINI E DENILSON DO AMARAL

 

This exploratory data analysis indicated the alternative use of asymetrical statistical
distribution for data analysis from time to flowering until seed maturation is achieved. It was used
data from the seed germination test, vigor tests, moisture content and seed dry weight as dependent
variables. The hybrids corn seeds were planted in three diferent times of seeding and samples of
plants were taken in each plot. The time interval of four days between each sample was used,
starting in the 23rd and finishing in the 59th. The Normal, Lognormal, Gumbel and Weibull
distribution was used to determine the highest numerical value of the loglikelihood as a test of
goodness of fit. The highest numerical value of the loglikelihood was provided by the Gumbel
distribution fit to the proportional data set but the seed dry weight and water content had a best fit
to the Weibull distribution.



Patrocinadores