Respect out1.gpar and out2.gpar parameters in ggdesplot

[SchmidtPaul]

Summary

This PR fixes ggdesplot() to properly respect user-provided out1.gpar and out2.gpar parameters for customizing outline appearance. Previously, outline colors, line widths, and line types were hardcoded, completely ignoring these documented parameters.

Changes

• Extract col, lwd, and lty from out1.gpar and out2.gpar lists
• Apply these parameters to geom_tileborder() calls (lines 465-480 in R/ggdesplot.R)
• Fixed a bug where hardcoded values didn't match documented defaults (e.g., out1 default lwd=3 was being rendered as lwd=1.5)

Testing

Tested with agridat::besag.met dataset. All examples work as expected:

library(desplot)
library(agridat)

# Test 1: Default colors (black and yellow)
ggdesplot(besag.met, yield ~ col*row|county, 
          out1=rep, out2=block,
          main="Default: black out1, yellow out2")

# Test 2: Custom colors with different line widths
ggdesplot(besag.met, yield ~ col*row|county, 
          out1=rep, out2=block,
          out1.gpar=list(col="red", lwd=3),
          out2.gpar=list(col="blue", lwd=2),
          main="Custom colors: thick red out1, blue out2")

# Test 3: Custom line types
ggdesplot(besag.met, yield ~ col*row|county, 
          out1=rep, out2=block,
          out1.gpar=list(col="darkgreen", lwd=2, lty="dashed"),
          out2.gpar=list(col="orange", lwd=1.5, lty="dotted"),
          main="Line types: dashed green out1, dotted orange out2")

# Test 4: Match the documented example styling
ggdesplot(besag.met, yield ~ col*row|county, 
          out1=rep, out2=block,
          out1.gpar=list(col="black", lwd=3),
          out2.gpar=list(col="gray50", lwd=1, lty=1),
          main="Styled like lattice examples")

Benefits

Users can now customize outline appearance as documented
Brings ggplot2 version closer to feature parity with lattice version
Fixes inconsistency between documented defaults and actual rendering
Enables matching the styling of lattice-based desplot examples

[kwstat]

Thanks for giving ggplot some fixes.