the challenge here is how to set the line styles so that line that share a higher-level index are grouped somehow. some notes from slack:
essentially, i was imagining adding in thicker lines to denote group averages, and then those could be stacked, so you get linewidth=1 on level 1, linewidth=2 reflects level 1 averages, linewidth=3 reflects level 2 averages, etc.
and we could also make things less transparent as you went up levels, e.g. alpha = 1/(level +1) + 0.2
the line colors should be determined by the highest level index (i.e. all data and group averages that share a common higher-level index should have the same line color and line style)
any linestyle arguments should have length "number of unique highest-level indices"
the challenge here is how to set the line styles so that line that share a higher-level index are grouped somehow. some notes from slack:
essentially, i was imagining adding in thicker lines to denote group averages, and then those could be stacked, so you get linewidth=1 on level 1, linewidth=2 reflects level 1 averages, linewidth=3 reflects level 2 averages, etc.
and we could also make things less transparent as you went up levels, e.g.
alpha = 1/(level +1) + 0.2the line colors should be determined by the highest level index (i.e. all data and group averages that share a common higher-level index should have the same line color and line style)
any linestyle arguments should have length "number of unique highest-level indices"