Learner feedback - sticky notes

[theavanrossum]

Morning 1: Spreadsheets & intro to R

Good:

Content:

• Introduction for using function building and using vector
• examples to explain functions
• I always used length() to find out how many elements in my vector fit the condition. "sum" really help, thanks.
• The basic of R. Definition of vectors and objects, always I had problem with them :)
• finding values in vectors + understanding them
• excel best practices. finally (for me) getting started in R

Style:

• I enjoyed the interactive session: listening and doing exercises
• I really enjoyed the rhythm of the class. Being able to do the exercises. Understand that sum works because logical operator gives true/false results.
• I like that there are exercises to practice
• Was really good to follow the session and exercises, good speed (not too slow/fast)
• live learning on screen *good way of explaining things. Like the signal post-it system. Great to have hands on possibility. Great preparation.
• the pace is excellent. great that you always checking with us if all is ok. sticky notes are great.

General

• All good so far :)
• An interesting introduction to RStudio. Enjoyed it :)
• all was well explained!
• I generally liked the content of the course and the way it was presented
• nicely done: explaining, how orders/tasks/functions and how to use them. just a little much ;)
• very didactic and good explanations of the subject
• very nice intro into R
• I enjoyed learning a new language. great course!

Could be better:

• excel data a little bit boring :)
• could be a bit quicker for the basics
• great deal of new info. a little fast if you haven't used R before but for majority should be ok. takes time to get used to new tasks and find everything.
• more exercises to catch the logic of the basic skills before we continue - thx!
• a bit more practical example on the filtering of tabular data
• could not eat in computer lab :)
• the room is a little cold

[theavanrossum]

Day 1 - PM

Qs

• can we have the code that you've types during the lesson?

  Good

• was a great learning experience :)
• really nice progression from basic functions into table manipulation
• I enjoyed dplyr and tidyr package
• a lot of new and useful information :)
• yeahh! I learned lots of R processing
• all good!

Could be better

• first part regarding the managing of the da [sic.] download, save etc a little bit difficult

[FloHu]

Day 2 - Morning

Good

• Good pace, easy to understand
• Fast help by instructors
• Learning many tools and how to use the help files.
• Starting the day with a challenge.

Could be better

• Meaning of height and width arguments to jitter wasn't clear at first.
• Had difficulties to recall all the stuff from yesterday. Maybe lack of practise.
• Session for section 2.3 (grouped summaries) too long and confusing.

[tavareshugo]

Day 2 - Afternoon

Good

• Several people liked quality of course:
  • good approach, great course, dynamic session
  • Learned a lot, very useful, relevant material
  • challenging and productive
• visualising data and their exercises
• working with databases was useful

Could be better

• Several people felt overwhelmed with information (a bit of tiredness also?):
  • too much information to retain
  • leads to frustration - realise lots can be done, but can't remember how to
  • more days in the course
• one-off post-its:
  • Would like more intro to base R
  • Customising plotting a bit fast (confusion with etherpad copy-paste issues)
  • Factors a bit challenging
• Several people also thought database in particular was fast (last session of day):
  • would like a more realistic example, and more exercises

[theavanrossum]

Questions from day 2 pm & day 3 am (and some leftovers from prior sessions)

general

• what to do if I get stick and nothing seems to work...???!!
• what do we use "R projects" for?
• what is rmarkdown? (it was suggested to the learner)
• can I use as many libraries as I want or do they interfere with each other?
• do I have to put my original data in some excel file and convert to csv so that I can open in R? or can I type in my data somehow? (with more simple data like some measurements)

bigger topics

• rmarkdown
• can you define new functions? (material from SC http://swcarpentry.github.io/r-novice-gapminder/10-functions/index.html )
• for loops (vs grouping in dplyr) ("what is the difference between grouping and for loops? are for loops used in R?") (for loops in R: https://swcarpentry.github.io/r-novice-inflammation/15-supp-loops-in-depth/index.html)

dplyr/data.frames

• demo rbind() ["can data transformations with dplyr be done with different tables? i.e. 3 tables with results from 3 biological replicates then create a new table with the avg and sd of the 3 biol. replicates?"]
• demo rename() ["how to use our own names of columns on the graph if they are not in the data.frame (renaming). Can I add a vector of names and use this?"]
• demo distinct() ["How to avoid duplicates in the tbl? (what if there is a duplicate data in a tbl and the number of rows doesn't reflect the number of samples? )"]
• demo grep()? tidyr::separate()? paste()? -> ["how to search/parse tables"]
• how do i pull out the data from a cell in a table? (just the data)
• how to deal with time data (hours, minutes)

databases

• specify more which commands are from dplyr and rsqlite package ... like left_joined won't work unless you load the rsqlite
• joins with != names (?)

plotting

• put two plots together (e.g. cowplot)
• are there R packages which makes it easier to modify features of a plot, for instance with a graphical interface?
• how can we show uncertainty in a plot? show sd in graph?
• can we add significance (p values) to plots? CI? R2?
• heatmaps of (selected) gene expression data? any particular packages?

factors

• revisit factors: how do I enter my own orders in the graph?
• can we have specific colors always used for the same category? same color for multiple categories?

statistical Qs

• how to process data? remove outliers/artefacts?
• "how do I do statistics?" :)
• lm() vs gm()

addressed in ether pad

• how do I create a RNA Seq file/matrix from my raw data to get 2-fold change etc
• where could we access some cool data to play with?

[FloHu]

Sticky notes day 3 (morning)

Good

• Reviewing what was already done and then moving forward.
• Good explanations, useful topics
• Special thanks to Thea for her part! :)
• Using RNA seq data and showing all the stuff one can do with it.
• Calculating correlations.
• Good to see how many visualistions are available and playing around with the options.

Could be better

• Too much improvisation
  • stay consistent when typing functions
  • move step by step rather than adding new lines
  • should retype code instead of modifying it
• There should be more time for analysing complex data sets (as the RNA seq data contained replicates, time points, wt vs. mutant etc.)
• RNA-seq data: difference between norm and trans was not clear, what does it mean exactly, how was it generated?
• In the source code file: put more comments and headers.
• Too fast (1x), part on correlation was too fast (1x)
• Beginning of the morning not fast enough.

[theavanrossum]

From our student from Japan, who said that the workshop was absolutely worth travelling so far for and that she would investigate whether she can invite us to teach at her institute :)

[tavareshugo]

Day 3 afternoon

Good

• (11x) Very good overall, learned a lot, loved it, thank you so much
• (5x) great instructors: dynamic, patient, answering lots of questions
• (3x) last Q&A session - lots of tricks to take home (RMarkdown got one thumbs

Could be better

• (3x) PCA a bit hard
  • partially related to usage of autoplot() and not explaining what type of object is prcomp
• (3x) a lot to take in - want more time
• (1x) did not understand why/what DESeq2 is doing
  • perhaps need more context for the data?
  • also renaming norm_cts --> raw_cts might make it less confusing
• (1x) did not understand difference between matrix and tbl
• (1x) cover loops and functions
• (1x) joining tables is quite hard