In this case I'm assuming you your y is of shape [n_lines, y-values], which you can directly pass as the data to add_line_collection()

y is indeed of shape [n_lines, y_values], but if I directly pass it to add_line_collection, I don't have the x-values. In your picture the x axis is from 0 to 100, even though it should be from 0 to 10.

Examples: ...

Thanks for the clarification, that helps a little bit, and the gallery example also helps (although it has quite a lot of things going on). It would help if the gallery example was linked in the documentation.

I suppose my main issue is that going from [10, 200] to [10, 200, 2] when I have the x values in a 1D vector is kind of tedious, you have to go through broadcast_arrays or similar, and wrangle around with the axes.

This issue of x/y-values could be easily resolved by using xarray, which I am a big fan of, although I understand if you want to stay fairly low level and work with raw numpy arrays.

I don't have the x-values. In your picture the x axis is from 0 to 100, even though it should be from 0 to 10.

ah, you can create the array such that you repeat the x-values linspace n_lines times:

# shape [n_lines, n_datapoints)
x = np.repeat([np.linspace(0, 10, 101)], 10, axis=0)

# you can even have different x-values per line if required by creating an [n_lines, n_datapoints] array

# shape [n_lines, n_datapoints)
y = np.array([np.sin(x[i] * i) + 2 * i for i in range(10)])

# stack along a new dimension
data = np.dstack([x, y])  # shape is [10, 101, 2]

fig = fpl.Figure()

fig[0, 0].add_line_collection(data)

fig.show(maintain_aspect=False)

Thanks for the feedback, I think we can make this clearer by creating mini-guides specific to each graphic.

We have plans to parse other types of data objects, like pandas dataframes (xarray would also fall under this), but it will come later, see: #395