juliaplots / plotdocs.jl Goto Github PK

Documenter makes this easy. And it would give a one-off stop for people who wonder why their plots suddenly look different and what to do about it.

The documentation on advanced layouts is confusing, and I think it might even be outdated. From what I can piece together, the a{0.3w} syntax is a reference to some obsoleted way of writing arrays, which is why the first link under "Advanced layouts" points to a seemingly unrelated page in the base documentation? But it's not explained what the a and b in the example mean, or how the array dimensions translate into a grid.

It might be that I'm slow, but as it stands I can't use the documentation to figure out how to create arbitrary grid layouts. Could someone who knows consider making the documentation a bit clearer?

I was browsing the attributes page and I noticed that on dark mode, it is hard to view the table.

I noticed that the latest build did not deploy to gh-pages. The master builds have actually been failing for a while, so the failures are not related to #110, but to some MkDocs configuration problem.

However, with #110, deployment failures are no longer being caught by Travis, which is a separate problem. I will try to look into this soon.

The figures on the layout page are off, at least for sure the two last. https://docs.juliaplots.org/latest/layouts/

Plots.jl is missing a tutorial. Before installation and a full blown manual on the backends, it really needs a tutorial that walks through the ideas one-by-one and introduces them with pictures.

Fixed by #184

We already have the examples, that are intended to show functionality. This would be for showing off the package and giving inspiration and be based on aesthetics. Inspiration e.g. http://www.r-graph-gallery.com/
Also came up in https://discourse.julialang.org/t/where-is-actual-development-in-plotting/6224/96

Clicking a plot could reveal the code.

Hi, I wonder why Plots.jl is split from it's documentation. I haven't found any issue relating to this.
As I see this now, docs are 1 month behind main project. Is there a reason to keep that this way?

If not, is there a possibility to merge this with main project? My point is that it's harder to improve on docs when they are hidden here. Also I wish to improve on docs, and that might require creation of docstrings in source.

Searching for aspect or ratio does not return anything helpful about the aspect_ratio attribute.

The icon for tabs is missing. I believe it is because of the missing asset favicon

If JS is disabled in the browser, some of the content doesn't show up, e.g. the attribute tables. The site looks completely functional and complete, so a warning about allowing JS for the docs in a <noscript>-tag could be useful.

Moving out the documentation from their readme.md's. This would make the ecosystem more coherent.

It is nice to have more clear direct examples to show how to use different backends for plotting in the documentation: http://docs.juliaplots.org/latest/install/
For example:
After you added your desired backend use one of the following commands to choose that

gr() # for GR
pyplot() # for PyPlot

I am still confused about how to switch to a different backend.

1. When I view it on github, the cool new axis logo looks odd in my browser:

perhaps it's an SVG glitch for this page, or for Safari, who knows....

2. If it was renamed to logo.svg or logo.png, could it (would it?) replace the old batman logo in the docs? I think it would look good:

3. I had occasion to draw it the other day so in case you ever need Julia code to render it, I saved it as a gist.

The KW dict in file https://github.com/JuliaPlots/Plots.jl/blob/master/src/arg_desc.jl contains more attribute descriptions than the PlotDocs tables in the online WEB view of http://docs.juliaplots.org/latest/attributes/.
For example I searched for the attribute 'camera' which is not available in table for subplots.

Hello,
trying to read the online documentation of Plots I stumbled over a broken main menu.
Starting with
http://docs.juliaplots.org/latest/
one sees a 'burger' menu in the top left near the title 'Plots'. It shows up on the left after left mouse button click. See attached screen shot.

Items which do not need a sub-menu can be selected and are shown.
Items which want to show a sub-menu just do a refresh - a sign of an error - and nothing happens.

Cheers,
Thomas

When I tried to install the package from GitHub:

(@v1.4) pkg> add https://github.com/JuliaPlots/PlotDocs.jl

I got the following error:

ERROR: invalid git HEAD (reference 'refs/heads/master' not found)

Suggestions?

After #73 the tutorial for StatPlots is again correct (i.e. the commands run and give the correct plots) but the flow of the explanation is a bit lost, as plotting a DataFrame is no longer a user recipe. Maybe a different user recipe example would be better and the DataFrame syntax could have its own space?

cc: @ChrisRackauckas

should add to "input data"

There used to be a shapefile example under ecosystem, but it appears to have gone missing. Would be nice with an example with fill_z describing this as a "choropleth", as this seems to keep coming up.

xs = collect(1:100)
μs = log.(xs)
σs = rand(length(xs))
scatter(xs,μs,grid=false,yerror=(σs,σs),group=xs)

This code works perfect fine before I update (can't remember which version was using). Now it keeps reporting

MethodError: no method matching split_attribute(::Symbol, ::Array{Float64,1}, ::Array{Int64,1})
Closest candidates are:
  split_attribute(::Any, ::Any, ::AbstractArray, !Matched::Any) at C:\Users\JD\.julia\packages\RecipesPipeline\2pkja\src\api.jl:34
  split_attribute(::Any, ::Any, !Matched::Tuple, !Matched::Any) at C:\Users\JD\.julia\packages\RecipesPipeline\2pkja\src\api.jl:36
  split_attribute(!Matched::Plots.Plot, ::Any, !Matched::Plots.SeriesAnnotations, !Matched::Any) at C:\Users\JD\.julia\packages\Plots\V8QVi\src\pipeline.jl:52

However, it works if not using group

scatter(xs,μs,grid=false,yerror=(σs,σs))

or if only use one vector for error bar

scatter(xs,μs,grid=false,yerror=σs,group=xs)

Problems with the new version?

I'm not sure how feasible it is, especially if we want to do it for all the backends, but I've noticed that in JuliaDB the plots in the documentation are generated while building the documentation. So for example this page generates this page. I think it's a very nice way of making sure that the docs are always accurate. On the other hand getting all backends to work remotely may be a tough challenge, but the "backend independent" docs could potentially be done like this (and the default backend as well).

  InsecurePlatformWarning
/usr/local/lib/python2.7/dist-packages/pip/_vendor/requests/packages/urllib3/util/ssl_.py:122: InsecurePlatformWarning: A true SSLContext object is not available. This prevents urllib3 from configuring SSL appropriately and may cause certain SSL connections to fail. You can upgrade to a newer version of Python to solve this. For more information, see https://urllib3.readthedocs.io/en/latest/security.html#insecureplatformwarning.
  InsecurePlatformWarning
  Downloading https://files.pythonhosted.org/packages/03/11/8ab0881bb6fe76e22d636c8327e86c2738f0d96f0d2145fec7e4fa7de9df/mkdocs_material-3.0.3-py2.py3-none-any.whl (574kB)
Collecting mkdocs>=1 (from mkdocs-material)
  Could not find a version that satisfies the requirement mkdocs>=1 (from mkdocs-material) (from versions: 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9, 0.10, 0.11, 0.11.1, 0.12.0, 0.12.1, 0.12.2, 0.13.0, 0.13.1, 0.13.2, 0.13.3, 0.14.0, 0.15.0, 0.15.1, 0.15.2, 0.15.3, 0.16.0, 0.16.1, 0.16.2, 0.16.3, 0.17.0, 0.17.1, 0.17.2, 0.17.3, 0.17.4, 0.17.5)
No matching distribution found for mkdocs>=1 (from mkdocs-material)
You are using pip version 9.0.1, however version 18.0 is available.
You should consider upgrading via the 'pip install --upgrade pip' command.
ERROR: LoadError: failed process: Process(`pip install --user mkdocs-material`, ProcessExited(1)) [1]
Stacktrace:
 [1] error(::String, ::Base.Process, ::String, ::Int64, ::String) at ./error.jl:42
 [2] pipeline_error at ./process.jl:695 [inlined]
 [3] #run#505(::Bool, ::Function, ::Cmd) at ./process.jl:653
 [4] run at ./process.jl:651 [inlined]
 [5] pip(::String, ::Vararg{String,N} where N) at /home/travis/.julia/packages/Documenter/bK0Xg/src/Deps.jl:34
 [6] top-level scope at none:0
 [7] include at ./boot.jl:317 [inlined]
 [8] include_relative(::Module, ::String) at ./loading.jl:1038
 [9] include(::Module, ::String) at ./sysimg.jl:29
 [10] include(::String) at ./client.jl:388
 [11] top-level scope at none:0
in expression starting at /home/travis/build/JuliaPlots/PlotDocs.jl/docs/make.jl:5

Who knows about this - @ChrisRackauckas ?

For example, in the main page, the pencil icon points to https://github.com/JuliaPlots/Plots.jl/edit/master/docs/index.md, which is a 404.

That could also produce an API page with docstrings for all exported functions.

There is a copyright notice in the footer: "Copyright © 2015-2016 Thomas Breloff". The first minor issue here is that it is outdated. But since there is no mention of it, it also seems to suggest that the docs themselves are not under a permissive license, even though LICENSE states that the software is under MIT. This is a slight contradiction in spirit, I think. Is it intentional or are the docs intended to be licensed under a CC license?

LICENSE itself is also outdated and displays "copyright 2016". Maybe the copyright notice should be updated automatically when the site is built?

Just a suggestion. I like the images in supported attributes but I can search for specific attribute https://juliaplots.github.io/supported/

Also would be helpful to have all possible values for some common attributes.

Hello
being in the online Plots documentation at
http://docs.juliaplots.org/latest/attributes/#series
In the table, for some rows the contents of the right columns is clipped.
See attached file, row for 'markershape' as example

Either table too small or line wrap is not enabled as in other tables.

Cheers,
Thomas

Hi,

There seems to be a leftover diff from a previous merge right at the beginning of recipes.md.

I created #124 to fix it (hopefully I kept the right header)

Cheers,

The documentation of section "savefig/png" should be improved to reflect the information available in the source code, for example as comments and real examples. Good for a better learning curve of newcomers.

• list of supported output formats also with same note from source "not all backends support every format"
• which output formats are supported by every backend
• default output format

In addition, some information about the resolution used in the output format. I created some png files for my package using gr() backend. They all had a resolution of 3600 x 2400 pixels which is ways too large.
Being able to define the wanted resolution would be great and could reduce / simplify the related workflow.

Github has a field for a project description and a link, in the main page of the repo. http://juliaplots.github.io/ should be added there.

https://stackoverflow.com/questions/59449553/cannot-convert-dataframe-to-series-data-for-plotting-in-julia

In particular, it seems the part about the dataframes syntax is incorrect

from JuliaPlots/Plots.jl#1369

Getting Started contains way too much.
Most projects have just Install instructions and some basic example/tutorial.

It actually took me over a year to realize that there was an attributes page, because I never read the getting started, since i got started I think before their were docs.

I suggest the following structure for the side bar:

• Home
• Learning
• Contributing
• Ecosystem
• Getting Started
  • Installation
  • Basics
  • Tutorial
• Manual
  • Input Data
  • Output
  • Layouts
  • Attributes
  • Colors
  • Animations
  • Backends
  • Supported Attributes
• Advanced Topics
  • Internals
  • Recipes
• Examples
  • PyPlot
  • GR
  • PlotlyJS
  • PGFPlots
  • UnicodePlots
  • InspectDR

In short it is basically move Tutorial into "Getting Started",
and move everything else into "Manual",
and reorder things.
Delete 0.7.

If you go here: ```

Those links are both incorrect.

Given this update: JuliaPlots/Plots.jl#578, the docs for the UnicodePlots backend could be updated to explain this usage. The canvas' don't seem to match up with the names given on the UnicodePlots README so this might need to give a list of the possible values (and which ones only use ASCII)

Per a conversation I had with Alan, I think it would be really great to add example links to each of the attributes: http://docs.juliaplots.org/latest/generated/attributes_plot/

It looks like the info is being brought in from here:
https://github.com/JuliaPlots/Plots.jl/blob/bf85afe9d6b67923cc000acaa1bdb499a87949b1/src/arg_desc.jl so would it be best to add the links to the examples there?

I can start finding examples and linking them if this is the right approach to add this.

@mkborregaard generated an awesome graph of method calls for histogram here, and it would be nice if we could include that in the docs. We could also look into using D3Trees.jl and building interactive tree visualizations for some plot calls; the code used to generate this PDF is at https://gist.github.com/mkborregaard/81825c3d370bb4d8dbfe59c3b2ae4b33.

All images in the example files could be loslessly reduced in file-size if thrown through PNG and animate GIF optimisers. I'm offering to do this, but I want to make sure it's actually useful before doing so.

Here is a dropbox folder with a few examples and tests:

https://www.dropbox.com/sh/wd7k4ecr0xhon73/AAA8JljU27Et4wfLshAnxxBDa?dl=0

The two gifs were downloaded from the front page of the docs. The optimised verions were a quick pass through GIMP's "Optimise for GIF", which should not affect visual output. The Lorenz Attractor example in particular became smaller. I know 8,6 MiB vs 10,8 MiB seems like nothing these days, but it's the principle of it ;). Also, death by a thousand paper cuts and all that.

Two reasons why I'm asking first : first, almost all automated optimisation tools remove all metadata from the PNG. So that is any EXIF, XMP and IPTC data in there. I want to make sure that this is OK.

Second, the usefulness ofthis depends on the way these images are added to the documentation. If they are auto-generated from the source code whenever the documentation or JuliaPlots itself updates, it wouldn't make any sense because I'm not going to manually do this every time ;). Also, PNG optimisers especially take a kind of brute-force approach, and they do take quite a bit of CPU power and time. So it's really only worth it if there is a minimum number of downloads for each image before being replaced.

But if these are all files that were generated at some point, and if they don't change very often, I'm willing to fork this repo, throw all of the PNGs and GIFs in docs/src/examples/img through an optimiser, and send a pull-request - from my POV it's just a quick drag/drop and forget about it until it's done a few hours later.

PS: currently there is a pending PR for better FFMPEG encoding flags for GIFs which in theory should apply better GIF optimisations from the start, and produce equal or even better output results that what I can do (both in size and visual quality), so if that PR is merged we might want to first regenerate all the GIFs in the documentation.

Repost of this issue: JuliaPlots/Plots.jl#2021

There could be a page with Histograms (describing all the functionality there), heatmap/contour-style plots, lines-and-or-markers (perhaps), bars-pies-shapes, 3d plots?
This is by far the most commonly requested documentation item on gitter and elsewhere. Better documented legends is nr 2.

Ref: https://discourse.julialang.org/t/package-compatibility-caps/15301

Couldn't find in the manual information about LaTeX symbols and how to format in a pretty way all the labels, texts, etc

We should adopt the new workflow for documenter: https://discourse.julialang.org/t/psa-use-a-project-for-building-your-docs/14974

I was reading http://docs.juliaplots.org/latest/input_data/ and I noticed that the example plot in the "Columns are series" section does not match the code. The mismatch is

markercolors = [:green :orange :black :purple
               :red   :yellow :brown :white]

This value results in lines that alternate colors,

but the accompanying figure shows the following setting.

markercolors = [:green :orange :black :purple]

Which one is it supposed to be? I'm happy to put together a pull request to update it accordingly.

I don't remember the issue that raised this, but it prompted me to write most of the docstrings: JuliaPlots/Plots.jl#1012

As discussed on Gitter:

if you browse through https://evizero.github.io/Augmentor.jl/ and see figures just take a look the the "Edit on Github" to see how I do it.
all but the "Tutorials" are pretty much vanilla Documenter. The tutorials though are all custom stuff, generated markdown pages by "examples/*.jl" scripts. so unless you want to mess with Weave just ignore those
Christof Stocker
@Evizero
00:30
I actually use Plots to produce the image + gif on this page https://evizero.github.io/Augmentor.jl/gettingstarted/ . forgot about that
Michael Krabbe Borregaard
@mkborregaard
08:33
So you just include an @eval block that creates the image? Sweet.
Christof Stocker
@Evizero
11:00
pretty much yea, and then you include that image in the markdown as usual
you can also use an @example block for this and just end the lines that save the figure with # hide