Plots.jl: Inconsistent behavior among backends for `x` and `y` of different length

Created on 9 Oct 2017 · 10Comments · Source: JuliaPlots/Plots.jl

I spent almost an hour trying to debug an algorithm thinking that it had a bug when actually the plot function was doing some magic behind the scenes:

x = collect(1:200)
y = collect(1:100)

plot(x,y)

cycle

As you can see, the plot is cycled whenever the input vectors have different length. Is there a good rationale for doing this instead of throwing an error or warning the user?

bug

Source

juliohm

All 10 comments

I can reproduce it with pyplot() backend.
However, with gr() I am getting the following:

using Plots
gr()
x = collect(1:200)
y = collect(1:100)
plot(x,y)
Error showing value of type Plots.Plot{Plots.GRBackend}: ERROR: BoundsError: attempt to access 100-element Array{Int64,1} at index [1:200] Stacktrace: [1] throw_boundserror(::Array{Int64,1}, ::Tuple{UnitRange{Int64}}) at ./abstractarray.jl:433 [2] checkbounds at ./abstractarray.jl:362 [inlined] [3] getindex(::Array{Int64,1}, ::UnitRange{Int64}) at ./array.jl:526 [4] gr_display(::Plots.Subplot{Plots.GRBackend}, ::Measures.Length{:mm,Float64}, ::Measures.Length{:mm,Float64}, ::Array{Float64,1}) at /home/leniac/.julia/v0.6/Plots/src/backends/gr.jl:973 [5] gr_display(::Plots.Plot{Plots.GRBackend}) at /home/leniac/.julia/v0.6/Plots/src/backends/gr.jl:551 [6] _display(::Plots.Plot{Plots.GRBackend}) at /home/leniac/.julia/v0.6/Plots/src/backends/gr.jl:1326 [7] display(::Base.REPL.REPLDisplay{Base.REPL.LineEditREPL}, ::MIME{Symbol("text/plain")}, ::Plots.Plot{Plots.GRBackend}) at /home/leniac/.julia/v0.6/Plots/src/output.jl:149 [8] display(::Base.REPL.REPLDisplay{Base.REPL.LineEditREPL}, ::Plots.Plot{Plots.GRBackend}) at ./REPL.jl:125 [9] display(::Plots.Plot{Plots.GRBackend}) at ./multimedia.jl:194 [10] eval(::Module, ::Any) at ./boot.jl:235 [11] print_response(::Base.Terminals.TTYTerminal, ::Any, ::Void, ::Bool, ::Bool, ::Void) at ./REPL.jl:144 [12] print_response(::Base.REPL.LineEditREPL, ::Any, ::Void, ::Bool, ::Bool) at ./REPL.jl:129 [13] (::Base.REPL.#do_respond#16{Bool,Base.REPL.##26#36{Base.REPL.LineEditREPL,Base.REPL.REPLHistoryProvider},Base.REPL.LineEditREPL,Base.LineEdit.Prompt})(::Base.LineEdit.MIState, ::Base.AbstractIOBuffer{Array{UInt8,1}}, ::Bool) at ./REPL.jl:646

lnacquaroli on 9 Oct 2017

👍1

Thank you @lnacquaroli , it seems that this issue is PyPlot specific then.

juliohm on 9 Oct 2017

So it seems, although when using PyPlot (outside Plots) I get an empty plot with the following error:

plot(x,y)
ERROR: PyError (ccall(@pysym(:PyObject_Call), PyPtr, (PyPtr, PyPtr, PyPtr), o, arg, C_NULL)) <type 'exceptions.ValueError'> ValueError(u'x and y must have same first dimension, but have shapes (200,) and (100,)',) File "/usr/lib/python2.7/dist-packages/matplotlib/pyplot.py", line 3318, in plot ret = ax.plot(*args, **kwargs) File "/usr/lib/python2.7/dist-packages/matplotlib/__init__.py", line 1894, in inner return func(ax, *args, **kwargs) File "/usr/lib/python2.7/dist-packages/matplotlib/axes/_axes.py", line 1406, in plot for line in self._get_lines(*args, **kwargs): File "/usr/lib/python2.7/dist-packages/matplotlib/axes/_base.py", line 407, in _grab_next_args for seg in self._plot_args(remaining, kwargs): File "/usr/lib/python2.7/dist-packages/matplotlib/axes/_base.py", line 385, in _plot_args x, y = self._xy_from_xy(x, y) File "/usr/lib/python2.7/dist-packages/matplotlib/axes/_base.py", line 244, in _xy_from_xy "have shapes {} and {}".format(x.shape, y.shape)) ...

lnacquaroli on 9 Oct 2017

Yes, this is definitely some Plots.jl magic, not something from Matplotlib.

juliohm on 9 Oct 2017

Plots is intended to cycle input arguments, this allows you to do things like scatter(1:4, [1]). But it's a good question why the backends are inconsistent! Plotlyjs has a third behaviour - it plots (1:100, 1:100). I hope you're OK with my renaming of the issue.

mkborregaard on 9 Oct 2017

👍1

I think plotlyjs behavior might be the most tolerable world if one wants to prevent error messages without falling into misunderstanding like pyplot cycling. Cycling the data explicitly is preferred overall, IMO.

lnacquaroli on 9 Oct 2017

👎2 👍1

There are so many inconsistencies across backends, I think we can focus the work and efforts in the Makie ecosystem with new recipes and so on. Closing this for now.

juliohm on 4 Mar 2020

@juliohm you've closed a number of issues without posting reasons. Is that because the issues no longer exist, or do you just want to (like above) signal that they are low priority for you?

mkborregaard on 4 Mar 2020

Most of them were too old and I could hardly reproduce given how much the
projects evolved. So if something shows up after the refactoring in vizcon
people can open more recent issues. They were just polluting the issue
tracker.

On Wed, Mar 4, 2020, 07:56 Michael Krabbe Borregaard <
[email protected]> wrote:

@juliohm https://github.com/juliohm you've closed a number of issues
without posting reasons. Is that because the issues no longer exist, or do
you just want to (like above) signal that they are low priority for you?

—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
https://github.com/JuliaPlots/Plots.jl/issues/1151?email_source=notifications&email_token=AAZQW3IIMAOQPHNUNEGWUVLRFYXVBA5CNFSM4D6HZCTKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOENXKJSA#issuecomment-594453704,
or unsubscribe
https://github.com/notifications/unsubscribe-auth/AAZQW3JZ3U3NOQZGQOVQT4DRFYXVBANCNFSM4D6HZCTA
.