I was reading about persistence.js in the source code and came across this line:

* component ID and will not [?] on with components without an ID.

Many times when I search for some R plotly documentation, I come across a page with a broken shiny app, which essentially renders that page useless. I usually ignored it, but I realize now that it would have been much more productive to make note and report it! So I'll start now :)

The app here doesn't exist: https://plot.ly/r/shinyapp-linked-click/

My frustrated experience using gganimate as someone who doesn't do this kind of thing very often:

• I exported an image to .gif using anim_save()

• The image looked pretty grainy (text and lines), and was 480x480

• I wanted to make it larger, and higher resolution, so I looked at the docs for ?animate() and for ?anim_save()

• I see that the gifski_renderer() is the default, a

We could use some good vignettes of both simple and advanced use.

The "components" returned from ml_pca() are NULL

# example
sc <- spark_connect(master = "local")
iris_tbl <- sdf_copy_to(sc, iris, name = "iris_tbl", overwrite = TRUE)

pca <- iris_tbl %>%
  select(-Species) %>%
  ml_pca()

pca$components
NULL

R session information:

devtools::session_info()
Session info -----------------------------------------------------------

The documentation of the nudge options does not specify the units of these values. It would help new users to clarify this.

They are probably in the same units as the data values for these axes? But they might’ve been relative units for the plotted axis (e.g. 0.5 would nudge the position by half of the axis length), or even absolute units for the plot (like the units for many of the other opti

Under "Keeps factor by default levels now", the second chart shows 100% thing1. It seems like it should show 100% thing2, just like the chart immediately above. That's what is in the data.

add a matrix diagram plot type, as described here: https://bost.ocks.org/mike/miserables/

It doesn't benefit much from the interactive possibilities of D3, but it's a valid network visualization type, so to be complete we should consider it.

✔  checking examples (32s)
   Examples with CPU or elapsed time > 5s
                  user system elapsed
   hc_boost     13.712  0.027  13.743
   hc_colorAxis  5.029  0.579   4.617

It would be great if the zoom'ed in area could be shown on the right of the full plot as well as on the left. Some times it makes more sense to guide the reader to first consider the full plot and then the zoom, which is easier (for those of us who read left to right) if the full plot is to the left of the zoom.

I just had a request for the algorithm used by geom_encircle(). This is what I sent; I ought to flesh this out slightly and add it to the docs.

(1) finds the convex hull of the points
(2) if desired (based on the 'expand' argument), moves the points defining the convex hull outward away from the centroid
(3) constructs an 'xspline' through these points; the details of this (and a literature

I was reading https://usethis.r-lib.org/reference/pr_init.html which unless I missed it doesn't mention https://usethis.r-lib.org/articles/articles/pr-functions.html which I nearly missed, so I'd suggest adding a link.

Base R colors can be used to form palettes (i.e. c(orange, orange1, orange2, orange3, orange4)). Tian Zheng created a layout of all base R colors in alphabetic order.

Reload homepage or navigate to any page, the "Built with..." text at the bottom left corner jumps up and down

For detailed entries, it'd be nice to be able to add bulleted details, as is common for many CV entries, for example: https://drive.google.com/file/d/0B9DjC_ju-eVCQnl3NjJsUkN0LTA/view

Do you think this is possible? How would one code that into the tibble/dataframe? I imagine there would be some tinkering with the Latex template too.

A major advantage of moving to the package structure is that we can use pkgdown to create a slick website to replace the README. In addition to documenting the package functions, each major section of the README can be an individual vignette/article.

Before posting your issue, please search the existing issues (both "open" and "closed") to check that a similar feature has not already been requested.

I read some created issues

1. the issues jdblischak/workflowr#103 and jdblischak/workflowr#37 (comment)
2. the discussion in the issue https://github.com/jdblischak/workflowr

This is a suggested code or documentation change, improvement to the code, or feature request

The package is great works in most conditions (many thanks for this) but also makes lazy that I don't want to wrangle misread pdf pages. Hence, I'd like to request below.

Provide a parameter/method to specify the number of columns; start and end co-ordinates of each column so that table is extracted

Pretty sure that stringi, stringr and rematch2 can be in suggests. Either the gadget takes into account whether these functions are loaded or it should.