ggpmisc (Preview)

Improve the automatic positioning of labels generated by stat_poly_eq(), most notably when grouping is in effect and with facets with free scales.  The values passed as argument to the new parameters label.x.npc and label.y.npc are in normalized parent coordinates with fractional values relative to the total dimension of the x and y scales. Character strings “right”, “center” and “left” for x and “top”, “center”, and “bottom” for y are also recognized. The behaviour of parameters label.x and label.y remains unchanged and they override any arguments passed through label.x.npc and label.y.npc. However, passing absolutes coordinates through the old parameters will be rarely needed.

Add stat_fit_deviations() for easily highlighting residuals in plots of fitted models, and stat_fit_residuals() for easily plotting residuals in plots matching plots of “lm” fits plotted with stat_smooth() even with grouping or facets. At the moment these two stats support lm() fits only.

Add stat_fit_glance() which uses package broom::glance() to produce a one-row data frame summarizing the model fit object. This gives maximum flexibility in model function choice at the expense of very frequently having to set aesthetics explicitly.

Add preliminary version of stat_fit_augment() which uses broom::augment() to enhance the data frame received as input with variables related to a model fit. At the moment functionality is very limited.

NOTE: This version of the package is now available from the R and R-test repositories at this site.

Share with

Leave a Reply