[FIXED] Python StatsModels Time Series Decomposition Duplicate Plot

Issue

I am using a mixture of Pandas and StatsModels to plot a time series decomposition. I followed this answer but when I call plot() it seems to be plotting a duplicate.

My DataFrame looks like

My index looks like

but when I plot the decomposition I get this

Strangely, if I plot only an element of the decomposition, the duplication does not occur

Solution

Assign the result of res.plot() to something, e.g. fig = res.plot(). The result of that method is a matplotlib.figure. The duplicate is from it's _repr_html_, which the notebook renders since it's the last item in the cell. You can try it yourself with

fig, ax = plt.subplots()
ax.plot([1, 2], [1, 2])
fig

The second method returns a matplotlib axes, which doesn't have a _repr_html so you just see that text above the figure.

Answered By - TomAugspurger