Edits to be made to existing tutorials

[beverlytan]

This issue thread will consist of all the edits that need to be made to improve existing Coding Club tutorials!

[beverlytan]

MCMCglmm tutorial

• The graphs for the trace plots of the models (the caterpillar ones) all have this dark shadow towards the bottom. Replace the files with ones without those shadows (they could still be screencaps, just better screencaps).

[Completed]

[beverlytan]

Occurrence tutorial

• The functions to do with coordinate cleaner do not work. Looks like there have been changes made to these functions in the package so this needs updating.
• Maybe the package got updated or something like that, this is the package repo that might have more info

[Pull request submitted]

[beverlytan]

Use of ggplot in tutorials

• Currently in the various tutorials when we use ggplot, sometime there is an object created, e.g. graph <- ggplot()... and other times it's just ggplot()
• I think it would be better if we are consistently creating objects when making graphs
• Check through the tutorials and add objects wherever the code is just ggplot().
• Also important that it has to be (object_name <- ggplot()...) e.g. the whole thing has to be in brackets, otherwise the graph is made, but not visualised, because we haven't "called" the object in R. And there should be a little snippet of text in each tutorial after the first ggplot graph that goes like "can't see your graph? maybe you have forgotten to put the brackets around the ggplot code chunk. Want to save your graph? Here is an example code chunk with ggsave"

[Identified problematic tutorials on 6 Jan]

• Problematic tutorials r/e ggplot visualization

  • Coding etiquette
  • Time series
  • Machine learning
  • Qualitative data
  • Spatial data and maps

• Mostly ok, just need to add notes about importance of brackets

  • Data vis 1: supplement
  • Occurrence: supplement
  • Model design: supplement
  • Git for labs: one liner
  • Data vis 2
  • SEEC
  • EVENET
  • Loops and functions
  • Mapping populations
  • GEE
  • Linear models
  • Mixed linear models
  • Generalised linear models in stan
  • INLA

[Completed]

[beverlytan]

Surveys and feedback

• Check that the links to the feedback surveys in all tutorials lead to a survey on the right topic. I think sometimes the links are wrong, e.g. it’s the data vis tutorial, but the modelling feedback survey.

• Similarly, in the separate repos for the data for each tutorial, there is a link to the respective feedback, but it’s not always right.

• Some tutorials don’t have surveys - could you add them? If you go to surveymonkey.com and log in using gndaskalova@gmail.com and use the same password as for twitter (I think…), you can see all the past surveys and you can just duplicate a survey and only change the relevant words, the surveys are pretty brief, so shouldn’t take too long I hope.

• Need to create new surveys

  • Intro to GEE
    • https://www.surveymonkey.co.uk/r/VKMZHD3
  • INLA
    • https://www.surveymonkey.co.uk/r/VVGXKZG
  • Ordination
    • https://www.surveymonkey.co.uk/r/VYLPZ29
  • Analyzing qualitative data
    • https://www.surveymonkey.co.uk/r/VMRWSKQ
  • Python crash course
    • https://www.surveymonkey.co.uk/r/VN5XKPR
  • Intro to fortran
    • https://www.surveymonkey.co.uk/r/VG5BK9T
  • Iris python data viz
    • https://www.surveymonkey.co.uk/r/VH6XDVZ
  • Meta-analysis with mcmcglmm has no survey link
    • https://www.surveymonkey.co.uk/r/V537CQ5
  • Machine learning in R has no survey link
    • https://www.surveymonkey.co.uk/r/77YWPQL
  • Numpy Python has no survey link
    • https://www.surveymonkey.co.uk/r/795PKMV
  • Topic modelling in Python has no survey link
    • https://www.surveymonkey.co.uk/r/7C5N3QV

• Easy fix: change on tutorial markdown

  • Manipulation and visualisation of spatial / occurrence
    • Survey link: https://www.surveymonkey.com/r/JYDDZ8F
  • Spatial data and maps
    • Survey link: https://www.surveymonkey.co.uk/r/NMD3N5K

• Other problems

  • Make Github for labs clearer in survey
  • Typo in stan GLM survey
  • Intro to python cc by sa size
  • Python pandas cc by sa size

[Pull requests submitted 15/01]

[beverlytan]

Data visualization 2 tutorial

• Do you want me to change the code for predictions / visualisations such that it's using the ggpredict() that we introduced in the model design tutorial?
  • I think we should have both options - so keeping Haydn's more complicated code, which shows you more of the calculations and the stuff "behind the scenes"
  • But also add the ggpredict() code as you suggest, so essentially making that section be "here are two ways to do the same thing, you decide which one suits your workflow best.”

[Issue #155, waiting for advice]

[beverlytan]

Model design tutorial

• Update the image files in the tutorial to also reflect the new theme (default CC theme)

[Completed]

[beverlytan]

Other generic edits

• Ordination tutorial: Fix graphs, points, numbers and titles

• Data clustering tutorial: Remove ggmap and create the map using ggplot

[Pull requests submitted 19/01]

[gndaskalova]

On the twitter front @beverlytan , there is a new tutorial on machine learning by Sam that you could tweet about sometime over the next few weeks (no rush at all) - https://ourcodingclub.github.io/2018/10/15/machine-learning.html

I'll go fix those gigantic license images in the tutorial now, since they look quite ugly, my bad, that was me trying to "fix" things and then making them worse!

[Completed]

[gndaskalova]

Happy new year @beverlytan !

Great work on the data vis 2 tutorial! I was wondering if you could update the colours on the final plot too so that they are the same as the other ones you made earlier in the tutorial? No rush on it, just whenever.

Thanks!

[Completed]

[gndaskalova]

Hi @beverlytan ,

Apologies for behind a bit behind on tasks here, I'll catch up over the weekend hopefully.

Just got a comment "to bring up to date" this tutorial https://ourcodingclub.github.io/2017/11/11/popchange.html from the online feedback form, could you go through it to see if there is indeed anything that needs updating?

Thanks!

[Pull request submitted 19/01]

[gndaskalova]

Hey @beverlytan , me again!

Just noticed the resolution of this graph in the functions and loops tutorials is a bit low, could you please swap the image file with a better one and increase the size of the graph in the tutorial a bit as well?

Thanks very much!

[Completed]

[gndaskalova]

Hey @beverlytan

John just added to this tutorial https://ourcodingclub.github.io/2017/02/08/funandloops.html could you please tweet about it sometime next week?

And in general, for the next semester, feel free to tweet about previous tutorials, those that haven't received much attention in the past, etc. Maybe something like 1 tweet about the tutorial from the workshop that happened in that week, and one tweet about some other tutorial, as a rough guideline.

[Completed]

[gndaskalova]

Hey @beverlytan

Thank you very much for sorting out the surveys, that will help us with keeping the tutorials up to date (and working)!

I just remembered that the links to the surveys are also in the README files in the repositories that hold the data for each tutorial - those ones are less urgent to fix, but whenever you can, can you check them as well? Then our surveys would all be sorted out!

[Completed]