# Sold

The plots in lexicographical order tell the story of the analysis of determining the functional relationship of the response variable perseat based on the potential explanatory variable(s):

report.ps
report.pdf

psl.rawsold.RData

## perseat univariate analysis:

qqmath(~perseat): qqnorm plot of perseat

boxcox(perseat ~ 1): boxcox plot of perseat.

qqmath(~log2perseat): qqnorm plot of log base 2 transformation of perseat, which possesses variable name log2perseat.

## log2perseat vs. row conditioned on area:

xyplot(log2perseat ~ row | area, subset(psl.rawsold.df, area == club) on top): Scatter plot of log2perseat against row conditioned on area with the club areas all on the top row.

### Model 1 (mod1): log2perseat against explanatory variables area and row

mod1

xyplot(resmod1 ~ year | area, subset(psl.rawsold.df, area == club) on top): Residuals of model 1 against year given area with club level on top row in the plot.

### Model 2: log2perseat against explanatory variables area, row, and year*I[area == club]

mod2

xyplot(resmod2 ~ year | area, subset(psl.rawsold.df, area == club) on top): Residuals of model 2 against year given area with club level on top row in the plot.

qqmath(~resmod2): qqnorm plot of residuals of model 2.

qqmath(~resmod2 | area): qqnorm plot of residuals conditioned on area of model 2.

xyplot(root4resmod2 ~ fitvalmod2): Fourth root of the absolute value of the residuals of model 2 vs the fitted values of model 2.

xyplot(root4resmod2 ~ fitvalmod2 | area): Fourth root of the absolute value of the residuals of model 2 vs the fitted values of model 2 by area.

dotplot(tapply(resmod2, area, sd)): Standard Deviation (sd) of resmod2 by area dotplot.

### Model 3: log2perseat against explanatory variables area, row, year*I[area == club], and secindex*I[area == "90e2"]

mod3

xyplot(resmod3 ~ year | area): Residuals of model 3 against year by area.

xyplot(resmod3 ~ row | area): Residuals of model 3 against row by area.