juliacomputing / plotlylight.jl Goto Github PK

View Code? Open in Web Editor NEW

96.0 4.0 6.0 1.28 MB

Plotly.js via Julia made easy.

Home Page: https://juliacomputing.github.io/PlotlyLight.jl/

License: Other

Julia 100.00%

julia plotlyjs data-viz plotting

plotlylight.jl's Introduction

PlotlyLight

PlotlyLight is an ultra-lightweight interface for working with Plotly.js.

✨ Features

🚀 Fastest time-to-first-plot in Julia!
🌐 Use the Plotly.js Javascript documentation directly. No magic syntax: Just JSON3.write.
📂 Set deeply-nested keys easily, e.g. myplot.layout.xaxis.title.font.family = "Arial".
📊 The Same built-in themes as Plotly's python package.

🚀 Quickstart

using PlotlyLight

preset.template.plotly_dark!()  # Change template

p = plot(x = 1:20, y = cumsum(randn(20)), type="scatter", mode="lines+markers")  # Make plot

p.layout.title.text = "My Title!"  # Make changes

p  # `display(p)` to see the updated plot

plotlylight.jl's People

Contributors

Stargazers

Watchers

Forkers

zsz00 cfarm6 disberd empet andreaskroepelin asbisen

plotlylight.jl's Issues

Option to specify default browser to display PlotlyLight.Plot() output

First of all, thank you for the nice package!
Here is my question:
How can I specify the default browser to display PlotlyLight.Plot() output?

For PlotlyJS.jl this can be achieved via the following lines in startup.jl

# --- redefine default browser for PlotlyJS: ---------------------------------------------------------------------------
using PlotlyJS
import PlotlyJS: PlotlyBase.launch_browser
launch_browser(tmppath::String) = run(`cmd /c start msedge.exe $tmppath`)

What would be the equivalent for PlotlyLight?

Downloading Plotly.js error

This is new (wit Julia 1.9.0 and 1.9.2):

[ Info: PlotlyLight: Downloading Plotly.js v2.24.3                                                                                                                                                                                 
ERROR: LoadError: InitError: RequestError: HTTP/1.1 403 Forbidden while requesting https://cdn.plot.ly/plotly-2.24.3.min.js

fail to precompile on 1.6

Hi,
the package seems to fail precompiling in 1.6 due to this line:

PlotlyLight.jl/src/PlotlyLight.jl

Line 258 in 689daa0

(; data, layout, config) = o

that uses the object destructuring syntax that was introduced in 1.7.

Didn't test if other issues arise on 1.6 after fixing that

Plots in VSCode Jupyter notebook opens in browser instead of below cell

Hi,

I'm trying to use the example plot script, however I end up with the plot opening up in the browser instead of below the cell in Jupyter notebook in VSCode.

using PlotlyLight

p = Plot(x = 1:20, y = cumsum(randn(20)), type="scatter", mode="lines+markers")  # Make plot

display(p)

The output of that is:

Process(`'C:\WINDOWS\system32\cmd.exe' /c start '%USERPROFILE%\.julia\scratchspaces\ec354790-cf28-43e8-bb59-b484409b7bad\CobWeb\index.html'`, ProcessRunning)

Running the same code in REPL creates the plot in a window withing VSCode which is how I would expect it to behave.

I'm running on Windows:
Julia v1.9.0
VSCode v1.82.3
PlotlyLight v0.7.4

Is it possible to have the plots show up in new tabs when displaying from a notebook? As it is now each new plot overwrites existing. I am on Apple silicon and running from VSCode notebook interface, if that matters.

VSCode Notebook - opens browser instead of cell rendering

If I run the Quickstart example from your README in a VSCode notebook directly (ipynb), it opens a html file in the local webbrowser instead of rendering it inside the cell as output (file:///home/XXXXXX/.julia/scratchspaces/ec354790-cf28-43e8-bb59-b484409b7bad/CobWeb/index.html).

Any idea why? I suppose this is not the expected behaviour? I would expect the chart to get rendered inside the notebook in the cell output by default without any special configuration.

Using VSCode 1.78.2 on Ubuntu 22.04 with extension Jupyter Notebook Renderers v1.0.15 installed. Julia 1.9.0.

Otherwise, love the idea of having a slim interface using the JavaScript API of Plotly! This should give the best performance and little maintenance required form the package side.

`SETTINGS.layout` gets mutated during plotting

It seems that sometimes the process of merging together a plot layout and the global layout is mutating the global layout.

To reproduce:

PlotlyLight.reset!()
Preset.PlotContainer.settings!(
    layout=Config(xaxis=Config(automargin=true)))
expected_layout = Config(xaxis=Config(automargin=true))
PlotlyLight.SETTINGS.layout == expected_layout # true
Plot(x=rand(10),y=rand(10),layout=Config(
    xaxis=Config(
        title=Config(
            text="myX"))))
PlotlyLight.SETTINGS.layout == expected_layout # no longer true

At the end, the SETTINGS.layout has acquired the xaxis.title.text field

Display in vscode

PlotlyLight.jl is great!

Can I display figures directly in vscode interactively instead of opening the browser?

feature: legendgrouptitle_text / legendgrouptitle.text not working

see: #28
You stated that legendgrouptitle_text is not included in plotly, but I can find a description here:
legendgrouptitle-text
I experimented also with legendgrouptitle.text

No plot shows up

On windows. PlotlyLight v0.6.2:

When I plot, a command window opens up with this message:

When I look at the mentioned file with a web browser, there is an actual plot:

I expected the plot to open in a browser by default.

Image Hosting

Missing NaN support

When trying to plot data containing NaN values, there's an error about those not being allowed in JSON spec

using PlotlyLight

p(x=0:11, y=[randn(5); NaN; randn(5)])

Error:

NaN not allowed to be written in JSON spec

Stacktrace:
  [1] error(s::String)
    @ Base .\error.jl:35
  [2] #write#90
    @ C:\Users\tbucelis\.julia\packages\JSON3\L8Yfy\src\write.jl:283 [inlined]
  [3] write
    @ C:\Users\tbucelis\.julia\packages\JSON3\L8Yfy\src\write.jl:282 [inlined]
  [4] write(::StructTypes.ArrayType, buf::Vector{UInt8}, pos::Int64, len::Int64, x::Vector{Float64}; kw::Base.Pairs{Symbol, Union{}, Tuple{}, NamedTuple{(), Tuple{}}})
    @ JSON3 C:\Users\tbucelis\.julia\packages\JSON3\L8Yfy\src\write.jl:187
  [5] write(::StructTypes.ArrayType, buf::Vector{UInt8}, pos::Int64, len::Int64, x::Vector{Float64})
    @ JSON3 C:\Users\tbucelis\.julia\packages\JSON3\L8Yfy\src\write.jl:181
  [6] write(::StructTypes.DictType, buf::Vector{UInt8}, pos::Int64, len::Int64, x::Config; kw::Base.Pairs{Symbol, Union{}, Tuple{}, NamedTuple{(), Tuple{}}})
    @ JSON3 C:\Users\tbucelis\.julia\packages\JSON3\L8Yfy\src\write.jl:157
  [7] write(::StructTypes.DictType, buf::Vector{UInt8}, pos::Int64, len::Int64, x::Config)
    @ JSON3 C:\Users\tbucelis\.julia\packages\JSON3\L8Yfy\src\write.jl:147
  [8] write(::StructTypes.ArrayType, buf::Vector{UInt8}, pos::Int64, len::Int64, x::Vector{Config}; kw::Base.Pairs{Symbol, Union{}, Tuple{}, NamedTuple{(), Tuple{}}})
    @ JSON3 C:\Users\tbucelis\.julia\packages\JSON3\L8Yfy\src\write.jl:187
  [9] write
    @ C:\Users\tbucelis\.julia\packages\JSON3\L8Yfy\src\write.jl:181 [inlined]
 [10] write(io::IOBuffer, obj::Vector{Config}; kw::Base.Pairs{Symbol, Union{}, Tuple{}, NamedTuple{(), Tuple{}}})
    @ JSON3 C:\Users\tbucelis\.julia\packages\JSON3\L8Yfy\src\write.jl:33
 [11] write
    @ C:\Users\tbucelis\.julia\packages\JSON3\L8Yfy\src\write.jl:30 [inlined]

Support nested NamedTuple

Instead of

p.layout.xaxis = (
    type = "log",
    dtick = log10(2),
    title = "xaxis"
)
p.layout.yaxis = (
    type = "log",
    dtick = 1,
    title = "yaxis"
)

Would it be possible to do

p.layout = (
    xaxis = (
        type = "log",
        dtick = log10(2),
        title = "xaxis"
    ),
    yaxis = (
        type = "log",
        dtick = 1,
        title = "yaxis"
    )
)

TagBot trigger issue

This issue is used to trigger TagBot; feel free to unsubscribe.

If you haven't already, you should update your TagBot.yml to include issue comment triggers.
Please see this post on Discourse for instructions and more details.

If you'd like for me to do this for you, comment TagBot fix on this issue.
I'll open a PR within a few hours, please be patient!

Creating stacked plots

It's not clear how to create stacks of plots:

https://plotly.com/javascript/subplots/

I tried passing a vector of plots, like:

    p1 = Plot(x=1:20, y=cumsum(randn(20)), type="scatter", mode="lines+markers")
    p2 = Plot(x=1:20, y=cumsum(randn(20)), type="scatter", mode="lines+markers")
    p = Plot([p1, p2], Config(; width=800, height=600))

but I get the error:

ERROR: MethodError: no method matching Plot(::Vector{Plot}, ::Config)

I next tried stacking the data:

    p1 = Plot(; x=1:20, y=cumsum(randn(20)), type="scatter", mode="lines+markers")
    p2 = Plot(; x=1:20, y=cumsum(randn(20)), type="scatter", mode="lines+markers")
    p = Plot(first.([p1.data, p2.data]), Config(; width=800, height=600))

But this resulted in both traces being placed in the same plot:

I next tried defining a grid in the config:

    p1 = Plot(; x=1:20, y=cumsum(randn(20)), type="scatter", mode="lines+markers")
    p2 = Plot(; x=1:20, y=cumsum(randn(20)), type="scatter", mode="lines+markers")
    p = Plot(
        [p1.data, p2.data],
        Config(; width=800, height=600, grid=(rows=2, columns=1, pattern="independent")),
    )

but this resulted in the second panel being left empty.

Any idea how I could set this up? It could be nice to have a Vector{Plot} function.

Documenter Docs

README is getting unruly and we could use some proper documentation.

Add Introductory resources to README

Not everyone knows Plotly.JS

There should be a link to a tutorials, etc. on how to use Plotly in the readme.

Is there some way to do animation a la `react!` of PlotlyJS?

UndefVarError: `Plot` not defined

Julia Version 1.6.7
Commit 3b76b25b64 (2022-07-19 15:11 UTC)
Platform Info:
  OS: Windows (x86_64-w64-mingw32)
  CPU: Intel(R) Core(TM) i7-10510U CPU @ 1.80GHz
  WORD_SIZE: 64
  LIBM: libopenlibm
  LLVM: libLLVM-11.0.1 (ORCJIT, skylake)
Environment:
  JULIA_SSL_CA_ROOTS_PATH =

Julia Version 1.9.0
Commit 8e63055292 (2023-05-07 11:25 UTC)
Platform Info:
  OS: Windows (x86_64-w64-mingw32)
  CPU: 8 × Intel(R) Core(TM) i7-10510U CPU @ 1.80GHz
  WORD_SIZE: 64
  LIBM: libopenlibm
  LLVM: libLLVM-14.0.6 (ORCJIT, skylake)
  Threads: 1 on 8 virtual cores
Environment:
  JULIA_SSL_CA_ROOTS_PATH =

The following fails on: JuliaLTS, v1.8.5 & v1.9.0:

using PlotlyLight
# Change template
PlotlyLight.template!("plotly_dark")
# Make plot
p = Plot(x = 1:20, y = cumsum(randn(20)), type="scatter", mode="lines+markers"); # this fails
# --- this works fine:
# p = PlotlyLight.Plot(x = 1:20, y = cumsum(randn(20)), type="scatter", mode="lines+markers");

The error message is:

WARNING: both PlotlyLight and PlotlyJS export "Plot"; uses of it in module Main must be qualified
ERROR: LoadError: UndefVarError: `Plot` not defined
Stacktrace:
 [1] top-level scope
   @ C:\data\git_repos\hycenta_stefan_pofahl\julia_scripts\examples\plotting\PlotlyLight\example_HelloWorld_PlotlyLight.jl:7
 [2] include(fname::String)
   @ Base.MainInclude .\client.jl:478
 [3] top-level scope
   @ REPL[1]:1
in expression starting at C:\data\git_repos\hycenta_stefan_pofahl\julia_scripts\examples\plotting\PlotlyLight\example_HelloWorld_PlotlyLight.jl:7

Jupyter Notebook blank

JSON Validation

Use the Plotly JSON chart schema to check a user's input into a plot.

Dark theme issue

Working around #4, with display, I am able to plot :)

However plotly_dark seems to not work

What am I doing wrong?

Save static figure

https://github.com/JuliaPlots/PlotlyJS.jl/blob/master/src/kaleido.jl

Issues with traces types that accepts matrices as inputs

Some plotly trace types accepts input data in matrix format and the behavior of the trace changes whether the data is specified as a matrix or as a play 1D vector.

The issue comes from the serialization of JSON3 that by defaults serializes any ND-Array as its 1D content.

This can create issues that are different depending on the specific trace affected. I could identify the following problems in these traces, with some details in the examples below:

heatmap
surface
contour

Heatmap Example

The heatmap trace can accept the z values as a matrix, in which case the x and y values can be provided as the vectors generating the underlying grid.
If z is provided as a vector, x and y also has to be provided as 1D vectors containing the x/y values of each single point in z.
This creates an issue when trying to provide z as matrix and x,y as generating vectors in PlolyLight:

Surface Example

A different problem happens in surface still when trying to provide x and y as generating vectors, where no Plot is produced with PlotlyLight:

Contour Example

Finally, something very weird that I don't understand happens with contour. The plot is rendered but the resulting contour is not the correct one:

Cannot display output file "index.html" on Ubuntu 22.04

see:
#28

Can't generate plot in Documenter.jl

When I use this package in an @example block, no output is rendered.

I would prefer this package over Plots.jl since it saves Github Actions time.

Can it run in Android's Termux's Ubuntu?

(@v1.7) pkg> test PlotlyLight
...
Progress [====================================  ✗ DefaultApplication
  ✗ Cobweb
  ✗ PlotlyLight
  0 dependencies successfully precompiled in 21 seconds (9 already precompiled)
  3 dependencies errored. To see a full report either run `import Pkg; Pkg.precompile()` or load the packages
     Testing Running tests...
ERROR: LoadError: IOError: stat("/proc/sys/kernel/osrelease"): permission denied (EACCES)
Stacktrace:
  [1] uv_error
    @ ./libuv.jl:97 [inlined]
  [2] stat(path::String)
    @ Base.Filesystem ./stat.jl:152
  [3] isfile(path::String)
    @ Base.Filesystem ./stat.jl:456
  [4] top-level scope
    @ ~/.julia/packages/DefaultApplication/7wyMn/src/DefaultApplication.jl:8
  [5] include
    @ ./Base.jl:418 [inlined]
  [6] include_package_for_output(pkg::Base.PkgId, input::String, depot_path::Vector{String}, dl_load_path::Vector{String}, load_path::Vector{String}, concrete_deps::Vector{Pair{Base.PkgId, UInt64}}, source::String)
    @ Base ./loading.jl:1318
  [7] top-level scope
    @ none:1
  [8] eval
    @ ./boot.jl:373 [inlined]
  [9] eval(x::Expr)
    @ Base.MainInclude ./client.jl:453
 [10] top-level scope
    @ none:1
in expression starting at /root/.julia/packages/DefaultApplication/7wyMn/src/DefaultApplication.jl:1
ERROR: LoadError: Failed to precompile DefaultApplication [3f0dd361-4fe0-5fc6-8523-80b14ec94d85] to /root/.julia/compiled/v1.7/DefaultApplication/jl_XtyYHr.
Stacktrace:
  [1] error(s::String)
    @ Base ./error.jl:33
  [2] compilecache(pkg::Base.PkgId, path::String, internal_stderr::IO, internal_stdout::IO, ignore_loaded_modules::Bool)
    @ Base ./loading.jl:1466
  [3] compilecache(pkg::Base.PkgId, path::String)
    @ Base ./loading.jl:1410
  [4] _require(pkg::Base.PkgId)
    @ Base ./loading.jl:1120
  [5] require(uuidkey::Base.PkgId)
    @ Base ./loading.jl:1013
  [6] require(into::Module, mod::Symbol)
    @ Base ./loading.jl:997
  [7] include
    @ ./Base.jl:418 [inlined]
  [8] include_package_for_output(pkg::Base.PkgId, input::String, depot_path::Vector{String}, dl_load_path::Vector{String}, load_path::Vector{String}, concrete_deps::Vector{Pair{Base.PkgId, UInt64}}, source::String)
    @ Base ./loading.jl:1318
  [9] top-level scope
    @ none:1
 [10] eval
    @ ./boot.jl:373 [inlined]
 [11] eval(x::Expr)
    @ Base.MainInclude ./client.jl:453
 [12] top-level scope
    @ none:1
in expression starting at /root/.julia/packages/Cobweb/AR1Ub/src/Cobweb.jl:1
ERROR: LoadError: Failed to precompile Cobweb [ec354790-cf28-43e8-bb59-b484409b7bad] to /root/.julia/compiled/v1.7/Cobweb/jl_xar3DG.
Stacktrace:
  [1] error(s::String)
    @ Base ./error.jl:33
  [2] compilecache(pkg::Base.PkgId, path::String, internal_stderr::IO, internal_stdout::IO, ignore_loaded_modules::Bool)
    @ Base ./loading.jl:1466
  [3] compilecache(pkg::Base.PkgId, path::String)
    @ Base ./loading.jl:1410
  [4] _require(pkg::Base.PkgId)
    @ Base ./loading.jl:1120
  [5] require(uuidkey::Base.PkgId)
    @ Base ./loading.jl:1013
  [6] require(into::Module, mod::Symbol)
    @ Base ./loading.jl:997
  [7] include
    @ ./Base.jl:418 [inlined]
  [8] include_package_for_output(pkg::Base.PkgId, input::String, depot_path::Vector{String}, dl_load_path::Vector{String}, load_path::Vector{String}, concrete_deps::Vector{Pair{Base.PkgId, UInt64}}, source::String)
    @ Base ./loading.jl:1318
  [9] top-level scope
    @ none:1
 [10] eval
    @ ./boot.jl:373 [inlined]
 [11] eval(x::Expr)
    @ Base.MainInclude ./client.jl:453
 [12] top-level scope
    @ none:1
in expression starting at /root/.julia/packages/PlotlyLight/xV2L1/src/PlotlyLight.jl:1
ERROR: LoadError: Failed to precompile PlotlyLight [ca7969ec-10b3-423e-8d99-40f33abb42bf] to /root/.julia/compiled/v1.7/PlotlyLight/jl_5RKypZ.
Stacktrace:
 [1] error(s::String)
   @ Base ./error.jl:33
 [2] compilecache(pkg::Base.PkgId, path::String, internal_stderr::IO, internal_stdout::IO, ignore_loaded_modules::Bool)
   @ Base ./loading.jl:1466
 [3] compilecache(pkg::Base.PkgId, path::String)
   @ Base ./loading.jl:1410
 [4] _require(pkg::Base.PkgId)
   @ Base ./loading.jl:1120
 [5] require(uuidkey::Base.PkgId)
   @ Base ./loading.jl:1013
 [6] require(into::Module, mod::Symbol)
   @ Base ./loading.jl:997
 [7] include(fname::String)
   @ Base.MainInclude ./client.jl:451
 [8] top-level scope
   @ none:6
in expression starting at /root/.julia/packages/PlotlyLight/xV2L1/test/runtests.jl:1
ERROR: Package PlotlyLight errored during testing

(@v1.7) pkg> st
      Status `~/.julia/environments/v1.7/Project.toml`
  [1b08a953] Dash v1.1.2
  [1b08a953] DashBootstrapComponents v1.1.0
  [682c06a0] JSON v0.21.3
  [9b87118b] PackageCompiler v2.0.6
  [f0f68f2c] PlotlyJS v0.18.8
  [ca7969ec] PlotlyLight v0.5.0
  [37e2e46d] LinearAlgebra

julia> versioninfo()
Julia Version 1.7.2
Commit bf53498635 (2022-02-06 15:21 UTC)
Platform Info:
  OS: Linux (aarch64-unknown-linux-gnu)
ERROR: IOError: uv_cpu_info: permission denied (EACCES)
Stacktrace:
 [1] uv_error
   @ ./libuv.jl:97 [inlined]
 [2] cpu_info()
   @ Base.Sys ./sysinfo.jl:216
 [3] versioninfo(io::Base.TTY; verbose::Bool)
   @ InteractiveUtils ~/julia-1.7.2/share/julia/stdlib/v1.7/InteractiveUtils/src/InteractiveUtils.jl:123
 [4] versioninfo(io::Base.TTY) (repeats 2 times)
   @ InteractiveUtils ~/julia-1.7.2/share/julia/stdlib/v1.7/InteractiveUtils/src/InteractiveUtils.jl:88
 [5] top-level scope
   @ REPL[9]:1

Usage of "legendgroup" inside PlotlyLight

Below a PlotlyJS example with one legendgroup per sub-plot.
Unfortunately, the code prevents to toggle visibility of single traces.
Do you know how arrange this functionality in PlotlyLight?

using PlotlyJS

function _legend_groups()
    # ---
    # --- create object containing sub-plots():
    _fig_obj = PlotlyJS.make_subplots(
        rows = 1, cols = 2,
    #           column 1    column 2
        specs = [ Spec()    Spec()], # raw 1
        )
    # --- read vector of standard colors "colorway":
    _colorway = PlotlyJS.PlotlyBase._get_colorway(_fig_obj.plot) # we must take the field ".plot"
    # ---
    _fig_layout = PlotlyJS.Layout(;
        title_text       = "Legend Groups",
        xaxis_title_text = "time / s",
        yaxis_title_text = "Amplitude",
        yaxis_range      = [0, 4],
        yaxis2 = PlotlyJS.attr(
            title       = "Amplitude",
            overlaying  = "y",
            side        = "left",
            range       = [0, 4],
            scaleanchor = "y",
            scaleratio  = 1,
        ),
        xaxis2_title_text = "time / s",
    )
    # ---
    for i_segment = 1 : 3
        if i_segment == 1
            PlotlyJS.add_trace!(_fig_obj, PlotlyJS.scatter(; y = i_segment .* ones(4), 
                line_color = _colorway[i_segment], 
                legendgroup = "group1", legendgrouptitle_text="First Group Title", ), row = 1, col =1)            
            PlotlyJS.add_trace!(_fig_obj, PlotlyJS.scatter(; y = 1.1 * i_segment .* ones(4), 
                line_color = _colorway[i_segment], 
                legendgroup = "group2", legendgrouptitle_text="Second Group Title", ), row = 1, col =2)            
        else
            PlotlyJS.add_trace!(_fig_obj, PlotlyJS.scatter(; y = i_segment .* ones(4), 
                line_color = _colorway[i_segment], legendgroup = "group1", ), row = 1, col =1)            
            PlotlyJS.add_trace!(_fig_obj, PlotlyJS.scatter(; y = 0.5 * i_segment .* ones(4), 
                line_color = _colorway[i_segment], legendgroup = "group2", ), row = 1, col = 2)
        end
    end
    # ---        
    PlotlyJS.relayout!(_fig_obj.plot, _fig_layout) # only re-layout the field ".plot"
    return _fig_obj
end

display(_legend_groups().plot)

Comonicon.jl (which uses PackageCompiler) builds ok but has a warning

https://github.com/KwatMDPhD/GSEA.jl v0.4.0 (there is the binary release with the warning).

You can try building gsea too to see the warning.

This may be due to deps/ path issue or something related? Because plotly.js should be downloaded into deps/?

Modifying attributes of a trace

Would it make sense to allow the plot to accept a Config argument? To accommodate for instance

t = PlotlyLight.Config(x = nths, y = y)
        t.name = "$(version)"
        t.mode = "markers+lines"
        t.marker = Config(symbol = "square", size = 12)
p(t) # instead of: push!(p.data, t)

plolty version

Hello!

Very cool package for minimizing TTF plot!
I have been playing around a lot with plotly and also started working recently on wrapping around PlotlyBase.jl to optimize plotting in Pluto with plolty (https://github.com/JuliaPluto/PlutoPlotly.jl)

I might consider using your package instead of PlotlyBase.jl as base, or even try to see if I can get to add some Pluto-specific optimizations as PR to this if the concept is OK with you and they are light enough!

Apart from this introduction, I was going through your code and noticed that at least for the CDN show option which is the default, you are pointing to plotly-latest for the plotly version.
I do not know if this was a deliberate choice to keep the library to version 1, but since plotly moved to version 2, one has to manually specify the version number and plolty-latest will always point to the last revision of version 1, as written here:
https://github.com/plotly/plotly.js#un-minified-versions-are-also-available-on-cdn

Please note that as of v2 the "plotly-latest" outputs (e.g. https://cdn.plot.ly/plotly-latest.min.js) will no longer be updated on the CDN, and will stay at the last v1 patch v1.58.5. Therefore, to use the CDN with plotly.js v2 and higher, you must specify an exact plotly.js version.

Would it be difficult to add some config parameter to also specify the plotly version to load? As quite a few fixes and features addition have happened since 1.58.5

Use latest release of plotly.js

Plotly.js recently released v2.22, which includes a new feature that I want to use. Is there a way to specify the version of plotly.js to use with this package?

stacked displays

Stacked displays do not seem to be prioritized (https://discourse.julialang.org/t/display-with-different-plots-show-the-same-plot-several-times/103897).
Related to JuliaLang/julia#30922 ?

referring to columns of a dataframe

When data refer to columns of a dataframe, life would be a little easier if no need to type the dataframe's name like in PlotlyBase. E.g.

# dfwithlongname::DataFrame has columns like n, open, high, low, and close etc.
# using PlotlyJS
# candlestick(dfwithlongname, x=:n, open=:open, high=:high, low=:low, close=:close, text=:dt)
using PlotlyLight
data = Config(x=dfwithlongname.n, open=dfwithlongname.open, high=dfwithlongname.high, low=dfwithlongname.low, close=dfwithlongname.close, text=dfwithlongname.dt, type="candlestick")

Thanks.

Unicode text isn't shown correctly with long series

Unicode text in axis title or legend isn't shown correctly with long series. For example, the yaxis title isn't rendered correctly with the code below.

Plot(
    Config(y=rand(20000)),
    Config(
        yaxis=Config(title="H₂ η ρ")
    )
)

I tried similar plots with Plotly.js directly and it worked, so it doesn't seem like an upstream issue.

Plotly.newPlot('graph', [{
  y: Array.from({length: 100000}, () => Math.random()),
}],{
  yaxis: {
    title: "H₂ η ρ"
  }
})

calling `JSON.json` on a PlotlyLight Plot causes stackoverflow

I stumbled upon this issues because of failing tests in PlotlyKaleido in the last days.

With version 0.8 of PlotlyLight, one can't seem to call JSON.json on the object without causing an error:

julia> using PlotlyLight, JSON

julia> p = Plot(Config(x = rand(10)));

julia> json(p)
ERROR: StackOverflowError:
Stacktrace:
  [1] Array
    @ .\boot.jl:477 [inlined]
  [2] Array
    @ .\boot.jl:486 [inlined]
  [3] Array
    @ .\boot.jl:494 [inlined]
  [4] _collect_indices
    @ .\array.jl:777 [inlined]
  [5] collect
    @ .\array.jl:761 [inlined]
  [6] JSON.Writer.CompositeTypeWrapper(x::Plot, syms::Vector{Any}) (repeats 74610 times)
    @ JSON.Writer C:\Users\Alberto.Mengali\.julia\packages\JSON\93Ea8\src\Writer.jl:22
  [7] JSON.Writer.CompositeTypeWrapper(x::Plot)
    @ JSON.Writer C:\Users\Alberto.Mengali\.julia\packages\JSON\93Ea8\src\Writer.jl:23
  [8] lower
    @ C:\Users\Alberto.Mengali\.julia\packages\JSON\93Ea8\src\Writer.jl:45 [inlined]
  [9] show_json
    @ C:\Users\Alberto.Mengali\.julia\packages\JSON\93Ea8\src\Writer.jl:348 [inlined]
 [10] #show_json#9
    @ C:\Users\Alberto.Mengali\.julia\packages\JSON\93Ea8\src\Writer.jl:359 [inlined]
 [11] show_json
    @ C:\Users\Alberto.Mengali\.julia\packages\JSON\93Ea8\src\Writer.jl:357 [inlined]
 [12] print(io::IOBuffer, obj::Plot)
    @ JSON.Writer C:\Users\Alberto.Mengali\.julia\packages\JSON\93Ea8\src\Writer.jl:383
 [13] sprint(f::Function, args::Plot; context::Nothing, sizehint::Int64)
    @ Base .\strings\io.jl:114
 [14] sprint
    @ .\strings\io.jl:107 [inlined]
 [15] json(a::Plot)
    @ JSON.Writer C:\Users\Alberto.Mengali\.julia\packages\JSON\93Ea8\src\Writer.jl:399

This does not happen with version 0.7 of PlotlyLight