@elizabethribble
In the Entering and Importing Data practical answers:

The answer to 1.1.20:

newmice3 <- read.csv(paste(mydir, "micedf1.csv", sep="/"), 
                     row.names = 1, stringsAsFactors = F)

using paste(mydir, is inconsistent with previous answers and raises questions why the answer is described this way. Leave it this way and accept the confusion and questions? Or change?

Apparently my changes to the ggplot practical didn't make it into the new reader or into github :P - David, will you please fix the issue from before (where aggregate creates two new columns with the same names?).

Answer to Question 1.2 is wrong, d is not local to f()

David will you please add your ggplot practical code to the repository? We need to fix the issue of the duplicate column names in the solution to the problem where you use the aggregate function. I think once that is updated we are all set to create the full course reader.

The aggregate example code in the question is not the same as the answer, the example code is still the old code with warnings.

In each, the following:

x=cut, y=color

...should instead be:

x=cutAgg, y=colorAgg

So, for example, for question 12:

ggplot(mean.price, aes(x=cut, y=color, fill=price)) +
geom_tile() + labs(title="Average prices")

Should instead be:

ggplot(mean.price, aes(x=cutAgg, y=colorAgg, fill=price)) +
geom_tile() +
labs(title="Average prices")

To explicit demonstrate the differences.

FYI for the next course: the ggplots2 lectures needs 90mins

The bookman font is not available on Mac, should change that in the final green theme plot.

@elizabethribble Comment from Martijn on this question; class_27, part B
Would make sense to indicate that an object or vector size is limited by the amount of available RAM. Would help with the understanding and motivation of teh question to indicate this and provide a clearer scope of the question.

In the lecture students are urged to correlate indepenant variables between each other and exclude those which show strong co-correlation.

1. This is not done in the prac ansers or commented on
2. what is a ball park r cutoff? 0.5?

Personally I think that the introduction about R is a bit too dragged out and we could use more time for the actual introduction of R code on the first day. Or just make it shorter to make the first day a bit less lecture heavy.

It looks like line 68 in 19_basic_course_on_R_linreg_practical_answers.Rnw is redundant. Shall we remove?

You sent me an email about practical question 11 in ggplot @elizabethribble . I've tested the code and it works, so I'm not sure what is wrong there.

Reminder to fix problem 11 in the ggplot practical (the names in the output of aggregate don’t match the names in the ggplot call).

Elizabeth

The line graph is supposed to be a dashed blue line, but the answer gives a solid blue line.

Martijn Kolijn: A 'live' genomics dataset workflow would add a lot of value towards the end of the course for a practical

@elizabethribble
In the Objects and Functions practical , 1.5 (e) the question:

"Use the function ``save()'' to save only the objects {\tt myFirstList} and {\tt pets} to a file called {\tt YourLastName_objects.RData} (replace YourLastName with your last name)."

myFirstList and pets were created earlier in the practical which is now commented.

My Q: can we bring back the part that was commented out to solve this? I propose it remains valuable!

In several practicals and lectures, stringsAsFactors = TRUE is assumed to be default, but this changed as of R 4.0.0. Everything needs to be updated to reflect this change.

Is it an idea to not keep the pdfs on git, but only the source files, to make sure that the most up to date version of the pdfs gets into the reader?
That would simply mean generating the pdfs before every iteration of the course, instead of copying the pdfs from the github.
Otherwise you might run into trouble when the pdfs and the source code do not match because edits have been made in the source, but the pdf has not been regenerated and posted yet.

Answers to problems 3 (m) and (n) in Hyp Test & CI 1 practical answers (pages 152 and 153 of May 2020 reader) have histograms that do not match what you get if you use the seed 1234 - need to update graphs

One of the Puromycin plots doesn't have a circle and a triangle in the legend, but two circles.

in section 4 of programming 1, writing a function, page 286 says "## note that z is returned" but that should say "note that z is not returned"

Code snippet to get started with:

set.seed(1234)
par(mfrow=c(2,2))
x <- sort(rnorm(10000))
plot(density(x))
plot(x)

## Help Description
## .Random.seed is an integer vector, containing the random number 
## generator (RNG) state for random number generation in R
runif(5, 1, 10)
runif(5, 1, 10)


## note 1: it only works on the FIRST subsequent call to R's (pseudo) random
## generator; after that it's de-activated
set.seed(123)
runif(5, 1, 10) # IMMEDIATELY run a fucntion that calls R's random generator 

## Re-run after the fact
runif(5, 1, 10)
runif(5, 1, 10)

## note 2: the number given determines the outcome of the random generator
set.seed(456)
runif(5, 1, 10)

Should this:
"are independent of each other"
Be:
"are dependent on each other"
?

In my experience, it is very confusing to many people why we use paste. I think we should remove it or give a clear example why and when you use paste to create (many) paths to load multiple files for instance.