The requirement to no longer specify source in the args dict tripped me (a layman) up when using a script that worked in 1.40 in 2.0.2. I couldn't work out what was wrong so I asked in a StackOverflow thread where @bryevdv ex

• I have searched the issues of this repository and believe that this is not a duplicate.

Reproduction link

Steps to reproduce

view the diagram in mobile or chorme opening devtool and click the icon of toggling

Color is listed as a parameter in the documentation for geom_tile/geom_rect/geom_raster, but is ignored for geom_raster. Based on the documentation and the behavior of other geoms, I would expect that color would be applied for geom_raster.

library(ggplot2)
df <- expand.grid(x = 0:5, y = 0:5)
df$z <- runif(nrow(df))
ggplot(df, aes(x, y, fill = z)) + geom_raster(color = "blac

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.

This can also demonstrate how they can be used with the new shiny vis_expect function from visdat.

I am building an interactive system that uses Jupyter widgets to control the settings that go into the Sankey diagram definition. The idea is to render the diagram when a button is pressed, at this stage I check the custom values entered by the user and adjust the diagram parameters accordingly.

I have the following proof of concept code, which is just a minor extension of the example from th

I suggest either adding a short code piece to use the rename() function to change the column "genus" to "genera" (thus alerting the learners to their relationship here, while adding a new function) or changing the column name in the original dataset. Otherwise, I've found that using the correct plural for genus confuses learners who are not biologists. Although it's the R ecology lesson and one

CodeCov looks like a good idea.

Seems really easy as we use Travis already. Are there other solutions?

We used to use emma, but I think that's been lost when we moved to gradle from ant.

I am sorry if I missed something in the docs. To know how to integrate a map in an RMarkdown document, I had to look at the source of vignettes. Is the solution used in vignettes the simplest one? Then, is it documented somewhere? And last, could it be simplified? E.g. in order to be able to use maps like magick objects.

Bower is now deprecated, and needs to be removed. All bower libraries should now be installed with npm, and statically referenced

The text uses two different articles when referring to files with the .R extension (i.e. "an .R file" and "a .R file"). This should be standardized. When referring to the file extension in speech, the ".R" is generally spoken as "dot are", so using the article "a" seems most appropriate.

The help pane could use additional calling out; the section describing mathematical functions (sin/cos/t

Dear Community,

There is a typo in the section titled "The StringsAsFactors argument" after the second block of code that demonstrates the use of the str() function. Right after the code boxes is written "We can see that the $Color and $State columns are factors and $Speed is a numeric column", but the box shows that the $Color column is a vector of strings.

Regards,

Rodolfo

In the Making Plots With plotnine lesson, there is a disclaimer that shows how to load plotnine.

I would recommend also mentioning here a reminder that plotnine is not in the standard Anaconda package and needs to be installed separately.

In this lesson, we learn that you can use a slice to get a substring and that "Numbers are not stored in the written representation, so they can’t be treated like strings." However, what about adding a short follow up to this part of the assignment to teach the students how to convert an integer to a string? For example, if you type in the code:

a = str(123)
print(a[1])

instead of the code

• Basic use case with QFU change and go-around
• Algorithms: particle model for wind field here
• Basic introduction on Enhanced Mode S (with link to https://mode-s.org/)
• Command line tool

For large datasets where computing the summary may be expensive, it would be useful to compute only part of it, be able to explore it, and then compute other parts of it without recomputing the initial report.

The selection of which parts to compute could be by:

• columns in the dataset,
• metrics, or
• row ranges.

In episode _episodes_rmd/12-time-series-raster.Rmd

There is a big chunk of code that can probably be made to look nicer via dplyr:

# Plot RGB data for Julian day 133
 RGB_133 <- stack("data/NEON-DS-Landsat-NDVI/HARV/2011/RGB/133_HARV_landRGB.tif")
 RGB_133_df <- raster::as.data.frame(RGB_133, xy = TRUE)
 quantiles = c(0.02, 0.98)
 r <- quantile(RGB_133_df$X133_HARV_landRGB.1, q

We should add headers to API requests so the client doesn't download things that haven't changed.

We should provide 304 when the resource hasn't changed.

Also, see it using etags makes sense.

https://thoughtbot.com/blog/introduction-to-conditional-http-caching-with-rails
https://thoughtbot.com/blog/take-control-of-your-http-caching-in-rails
https://medium.com/@jatescher/basic-rails-ap

Have you taught this lesson? One way you can help us improve it is by filing issues here about typos you've discovered, but you can also help us fill out the instructor notes!

• What did you find difficult?
• Where was it too fast?
• What exercises could be improved?
• What did learners struggle with?
• What did you emphasize? W