I've used R's ggplot for ages and its API is simply wonderful and intuitive. Matplotlib's API is always unpleasant in comparison. Plotly's API is better but ggplot still wins. Obviously matplotlib's depth and breadth are tremendous, but that API isn't fun or memorable or particurlarly orthogonal.

Plotnine is quite close to ggplot so my learning curve was negligible. I get the impression plotnine's adoption hasn't been huge and I'm wondering why. Is there a lack of beginner material, targeted to python/pandas users? There's lots of introductory ggplot material but it's all R and assumes dplyr/tidy knowledge, which would all be foreign to python/pandas people.