Multiple Plots Layout Weird Behavior

I have created a plot using multiple curves. Individual plots work fine. But when I try to display side by side it just mixes everything.

def get_line_chart(plot_data, agg, agg_by, variable_col, direction):
    curves_list = []
    xaix = sorted(plot_data[agg].dropna().drop_duplicates())
    
    for turnDirection, data in plot_data.groupby([variable_col]):
        curve = hv.Curve((data[agg], data[agg_by]), name=str(turnDirection), label=str(turnDirection)) \
            .opts(xlabel=agg, xrotation=90, ylabel=agg_by)
        curves_list.append(curve)

    from functools import reduce
    return reduce(lambda x, y: x * y, curves_list).opts(title=direction, width=700)

def create_plot(df1, maxIntersections, agg_present, agg_by, variable_col, outputFolder):
    plots = []
    for direction in ['East', 'West',  'North', 'South']:
        filtered_data = df1[df1["inboundDir"] == direction]
        plots.append(get_line_chart(filtered_data, agg_present[1], agg_by, variable_col, direction))
    from functools import reduce
    plot = hv.Layout(reduce(lambda x, y: x + y, plots)).cols(2)
    hv.save(plot, outputFolder + 'WorstIntersectionsAreaPlot{0}.html'.format(variable_col))
       

This is normal NORTH plot

This is normal SOUTH plot

When I try to merge in a single view using + operator suddenly the second plot mixed up.

If I add multiple plots, 1st one behaves normally, all others are mixed up.

[philippjfr]

I’d also expect that using hvPlot all of your code reduces to some variation of:

df1.hvplot.line('timeBin', agg_by, by=variable_col, groupby="inboundDir", shared_axes=False).layout().cols(2)

[philippjfr]

As a separate note, this bit can be simplified from:

    from functools import reduce
    return reduce(lambda x, y: x * y, curves_list).opts(title=direction, width=700)

to:

    return hv.Overlay(curves_list).opts(title=direction, width=700)

and similarly you can just do:

plot = hv.Layout(plots).cols(2)