scale differs when starting at zero about plotnine

[votti]

I think this is the 'correct' or at least the behaviour consistent with R ggplot2. coord_cartesian (not scale_y_continous) should be used instead to not ignore the data outside the provided limits, but rather just 'zoom in' the axes:
http://ggplot2.tidyverse.org/reference/coord_cartesian.html

[aflueckiger]

@votti Thanks for pointing out that I used the wrong statement to adjust the visible scale. However, the issue is still present even when using the coord_cartesian syntax.
This code produces the plot below that doesn't start at zero yet.

plot = (ggplot(df_docs_model) +
    aes(x='quarter_num', y='value', color='factor(topic)') +
    geom_smooth(method='loess', span = 0.1, alpha=0.2, size=2, show_legend=False) + #  
    scale_x_continuous(breaks=list(range(1994, 2019, 2))) +
    scale_y_continuous(breaks=[y / 100 for y in range(0, 15, 1)]) +
    coord_cartesian(ylim = (0, 0.14)) +
    facet_wrap('~topic'))

Is this feature already properly implemented in plotnine? I'm asking this again since the respective ggplot2 documentation, posted by votti, explicitly specifies this use case.

[has2k1]

@votti is right. However, there was a bug in coord_cartesian when setting the ylim. It was fixed in #30.

Try the development version

pip install git+https://github.com/has2k1/plotnine.git

to see if it works for you.

[aflueckiger]

It worked fine with the current development version. I appreciate your help. Thanks.