Information for the 2019 Applied Machine Learning workshop

workshop

#1

We will be sending out emails with the official information regarding the Applied Machine Learning workshop. This post will be used to supplement that information and enable a discussion about prerequisites, package installs, and any other questions that you have beforehand.

The GitHub repo rstudio-conf-2019 will contain the code, slides, data, and everything else required for the workshop. It will be finalized a few days before the we start.

Logistics

There will be about 180 people taking this two-day workshop. To accommodate everyone, there will be two parallel sessions running next door to one another. I will be teaching blocks of the materials in each room and switching back and forth. When I am not in a session, Alex Hayes or Davis Vaughan will teach the materials. There will also be multiple teaching assistants in each room to help with installs and hands-on sessions.

The logistics might evolve a little before the conference. We'll maintain contact over email to keep you posted and update this Community post with anything that you need to know.

Software

We will have RStudio Server Pro (RSP) instances for those would like them. If you don’t want to use a server (i.e. you will only run on your machine), the installation instructions are contained in installs.R. We recommend the current release of R for compatibility. If you don't want to modify your local install to upgrade, we will be happy to provide an RSP instance for you to use.

We will provide internet access in the room. Your browser will need to be able to use javascript to see the HTML slides.

Discussion

If you have questions, please post them here so that everyone can be informed and contribute to the discussion.

Thanks and see you in Austin.


rstudio::conf Workshop Information
#2

#3

#4

Thanks Max. I ran the installs.R. I'm running R version 3.5.2.
I get an error when loading the tidymodels library - it doesn't like my tibble version.

#> try loading tidymodels individually
library(tidymodels)

Error: package or namespace load failed for ‘tidymodels’ in loadNamespace(j <- i[[1L]], c(lib.loc, .libPaths()), versionCheck = vI[[j]]):
 namespace ‘tibble’ 1.4.2 is already loaded, but >= 2.0.0 is required

What version of tibble do I have?

packageVersion("tibble")
[1] ‘1.4.2’

It is indeed 1.4.2. How do I get version >= 2.0.0?
What if I update it?

update.packages("tibble")
packageVersion("tibble")
[1] ‘1.4.2’

Nope. Try installing it directly.

install.packages("tibble")
Installing package into ‘blah blah’
(as ‘lib’ is unspecified)

  There is a binary version available but the source version is later:
       binary source needs_compilation
tibble  1.4.2  2.0.0              TRUE

  Binaries will be installed
trying URL 'https://cran.rstudio.com/bin/windows/contrib/3.5/tibble_1.4.2.zip'
Content type 'application/zip' length 243706 bytes (237 KB)
downloaded 237 KB

package ‘tibble’ successfully unpacked and MD5 sums checked
Warning in install.packages :
  cannot remove prior installation of package ‘tibble’

The downloaded binary packages are in 'blah blah'

It can't remove prior installation? Is that a problem? Restart RStudio, load and look at the version again.

library(tibble)
packageVersion("tibble")
[1] ‘1.4.2’

Bananas. Any suggestions?


#5

I encountered this fairly recently. Perhaps you might still have the old version attached.

What happens if you restart a new R session by hitting:

ctrl + shift + F10

and then:

install.packages("tibble")

#6

(I'm assuming that you are on Windows)

It's just bad timing. A new version of rsample was sent in that needed changes for the new version of tibble. Even though tibble was approved a few days before, the rsample binary is available first.

So wait a few days and try again. You can check this page first to see if the binary has been built and mirrored. Right now it has:

Windows binaries:
r-devel: tibble_1.4.2.zip, r-release: tibble_1.4.2.zip, r-oldrel: tibble_2.0.0.zip

You'll want those to say 2.0.0 before trying to install again.

EDIT

There is an issue with tibble 2.0.0 and a patch is being sent to CRAN. Stay tuned and I'll let you know how to proceed.


#7

I encountered this error when running the script:

Error in session_info() : could not find function "session_info"

I assume the package devtools should be loaded before running the script?


#8

You can run

sessioninfo::session_info()

instead. I've updated the install file too.


#9

@Vanichols: the binaries for version 2.0.1 are available now.