Dataset:  hazelnut_color.csv

Source: A. Simsek (2007). "The Use of 3-D Nonlinear Regression in
Mathematic Modeling of Colour Change in Roasted Hazelnuts,"
Journal of Food Engineering, Vol. 78, pp. 1361-1370.

Description: Changes in L colour parameters for 3 varieties of
hazelnuts (Fosa, Palaz, Tombul) under Whole and Ground conditions
at various temperatures (C) and minutes of exposure.
Note that the control condition (Process=1) was included in 
1-Way ANOVAs comparing
the process conditions, but not in regression models.
The mean values were used in regressions as individual observations.
There were 3 replicates per process condition.

Regression Models: (Y=L change, X1=Temp, X2=Minutes)
Model 1: Y = B0 + B1*X1 + B2*X2 
Model 2: Y = B0 + B1*X1 + B2*X2 + B11*X1^2 + B22*X2^2
Model 3: Z = B0 * exp[-0.5 * (((X1-B01)/B1)^2 + ((X2-B02)/B2)^2)]
Model 4: Z = B0 / [[1+ ((X1-B01)/B1)^2] * [1 + ((X2-B02)/B2)^2)]]

Models 3 and 4 are called Gaussian and Lorentz regression models,
respectively.

Variable Names (W=whole, G=ground, mn=mean, sd=std. dev.):
Process
tempC
minutes
fosaWmn
fosaWsd
fosaGmn
fosaGsd
palazWmn
palazWsd
palazGmn
palazGsd
tombulWmn
tombulWsd
tombulGmn
tombulGsd