cwwhitney/decision_analysis_course

🎓 Uni-Bonn Decision Analysis graduate course, lectures and materials

decision_analysis_course's Issues

Module content file doe not match schedule

We need to update the overview pdf to reflect the schedules
See:
Added graphical syllabus

Multi-criteria Decision Making

So far, as far as I know, all the different possible decision criteria/objectives are combined into one number (usually NPV) to be able to come up with decision suggestions and calculate EVPI and so on.

It would be nice, to be able to report on the "raw" decision outcomes (might be CO2, money, a nutrition metric, ...), be explicit about trade-offs and then weight them in the very last step (possibly consulting decision-makers / stakeholders again). "Option A is more slightly climate-friendly, but option B is a lot more profitable, ... We decided for option X because of ..."

The HoliDA package will be built to support this kind of workflow. Maybe already conceptualize it this way in the course/book/paper?

Glossary of terms

Add a glossary of al the terms used and a definition of these as they apply to this research

cf. this from the EU

Unstable - Rstudio Connect and the Linux server - parse error: trailing garbage

This and other websites hosted on Rstudio Connect and the Linux server seem to be down:

The error is about 'parse error: trailing garbage'

1/27 10:29:17.210 (GMT)
Running on host: inresgb-0002
11/27 10:29:17.232 (GMT)
Linux distribution: Ubuntu 20.04.6 LTS (focal)
11/27 10:29:17.236 (GMT)
Using Packrat dir /opt/rstudio-connect/mnt/app/packrat/lib/x86_64-pc-linux-gnu/4.1.2
11/27 10:29:17.244 (GMT)
Server version: 2021.12.1
11/27 10:29:17.244 (GMT)
LANG: en_US.UTF-8
11/27 10:29:17.244 (GMT)
R version: 4.1.2
11/27 10:29:17.244 (GMT)
shiny version: 1.7.1
11/27 10:29:17.244 (GMT)
httpuv version: 1.6.3
11/27 10:29:17.244 (GMT)
rmarkdown version: 2.11
11/27 10:29:17.244 (GMT)
knitr version: 1.37
11/27 10:29:17.244 (GMT)
jsonlite version: 1.7.2
11/27 10:29:17.244 (GMT)
RJSONIO version: (none)
11/27 10:29:17.244 (GMT)
htmltools version: 0.5.2
11/27 10:29:17.244 (GMT)
Using pandoc: /opt/rstudio-connect/ext/pandoc/2.16
11/27 10:29:17.454 (GMT)
Using jsonlite for JSON processing
11/27 10:29:17.459 (GMT)
Using Shiny bookmarking base directory /opt/rstudio-connect/mnt/bookmarks
11/27 10:29:17.459 (GMT)
11/27 10:29:17.459 (GMT)
Starting R with process ID: '1218904'
11/27 10:29:17.508 (GMT)
Shiny R Markdown document starting ...
11/27 10:29:18.404 (GMT)
Error in value[3L] : parse error: trailing garbage
11/27 10:29:18.404 (GMT)
":{},"value":["0.10.1"]}]}]} {"type":"list","attributes":{},
11/27 10:29:18.404 (GMT)
(right here) ------^
11/27 10:29:18.404 (GMT)
Calls: local ... tryCatch -> tryCatchList -> tryCatchOne ->
11/27 10:29:18.404 (GMT)
Execution halted
11/27 10:29:18.404 (GMT)
Shiny R Markdown document exiting ...

Introduction: Making the overview cheat sheet more readable

The overview cheat sheet is a little small and at least on my laptop I can barely read it. Would be good if it is bigger or that you can select it to show it bigger or download it.

Add glossary

Many terms have vastly different meanings, depending on who you ask. So to be explicit, what we are talking about, Katja suggested adding a glossary..

Interesting, but ambiguous terms include:

• uncertainty
• risk
• decision criterion
• objective
• utility
• system
• holistic
• systems thinking
• impact pathway
• theory of change

json related trailing garbage

Error at https://agtools.app/decision_analysis/

Realted to jsonlite

It might been something wrong with rmarkdown:::shiny_prerendered_prerender(input_rmd, rendered_html, output_dir, prerender_option).

Error in value[3L] : parse error: trailing garbage
":{},"value":["0.10.1"]}]}]} {"type":"list","attributes":{},

Logs:

12/14 14:22:54.188 (GMT)
Running on host: inresgb-0002
12/14 14:22:54.210 (GMT)
Linux distribution: Ubuntu 20.04.6 LTS (focal)
12/14 14:22:54.217 (GMT)
Using Packrat dir /opt/rstudio-connect/mnt/app/packrat/lib/x86_64-pc-linux-gnu/4.1.2
12/14 14:22:54.226 (GMT)
Server version: 2021.12.1
12/14 14:22:54.226 (GMT)
LANG: en_US.UTF-8
12/14 14:22:54.226 (GMT)
R version: 4.1.2
12/14 14:22:54.226 (GMT)
shiny version: 1.7.1
12/14 14:22:54.226 (GMT)
httpuv version: 1.6.3
12/14 14:22:54.226 (GMT)
rmarkdown version: 2.11
12/14 14:22:54.226 (GMT)
knitr version: 1.37
12/14 14:22:54.226 (GMT)
jsonlite version: 1.7.2
12/14 14:22:54.226 (GMT)
RJSONIO version: (none)
12/14 14:22:54.226 (GMT)
htmltools version: 0.5.2
12/14 14:22:54.227 (GMT)
Using pandoc: /opt/rstudio-connect/ext/pandoc/2.16
12/14 14:22:54.445 (GMT)
Using jsonlite for JSON processing
12/14 14:22:54.449 (GMT)
Using Shiny bookmarking base directory /opt/rstudio-connect/mnt/bookmarks
12/14 14:22:54.449 (GMT)
12/14 14:22:54.449 (GMT)
Starting R with process ID: '1549229'
12/14 14:22:54.500 (GMT)
Shiny R Markdown document starting ...
12/14 14:22:55.415 (GMT)
Error in value[3L] : parse error: trailing garbage
12/14 14:22:55.415 (GMT)
":{},"value":["0.10.1"]}]}]} {"type":"list","attributes":{},
12/14 14:22:55.415 (GMT)
Shiny R Markdown document exiting ...
12/14 14:22:55.416 (GMT)
(right here) ------^
12/14 14:22:55.416 (GMT)
Calls: local ... tryCatch -> tryCatchList -> tryCatchOne ->
12/14 14:22:55.416 (GMT)
Execution halted

Intro slides and content

Add the 'Tropentag talk' overview of DA for agriculture.

Earlier conceptual model presentation

It seemed like the progress of some groups wasn't where we expected. An earlier presentation (by at least two weeks), followed by guidance might give them enough time to work on the coding and literature review.

Crashes on legeleistung in new Practical branch

The chunk legeleistung in the new 'Practical_Schiffers.Rmd' crashes my R session... Could not figure it out yet...

Minor sp package issues (deal with this later)

Note:

The legacy packages maptools, rgdal, and rgeos, underpinning the sp package,
which was just loaded, will retire in October 2023.
Please refer to R-spatial evolution reports for details, especially
https://r-spatial.org/r/2023/05/15/evolution4.html.
It may be desirable to make the sf package available;
package maintainers should consider adding sf to Suggests:.
The sp package is now running under evolution status 2
(status 2 uses the sf package in place of rgdal)

Schedule and reading assignments review

Review the schedules and assignments

Server connection unstable

The https://agtools.app/decision_analysis/ website is unstable. Crashing and reloading a lot

The static html page is working fine https://htmlpreview.github.io/?https://github.com/CWWhitney/Decision_Analysis_Course/blob/main/Index.html#section-model_programming

Error: Error: Internal Server Error [500]

Unable to chase this error - Everything on the Rstudio Connect server ran well and the page said 'Activation Complete' but the website just has 'Error: Error: Internal Server Error [500]'. No notes from the build configuration.

Find and link primary sources

I suspect (and in some cases with the EVPI know), that many of the things in How to Measure Anything originate from scientific literature. E. g. the definition of and method to calculate the EVPI are found in scientific literature from at least the 60s.

It can probably be a bit of work, but I would find it interesting to ideally cite the original sources of these concepts.

How to include other sources of data than expert estimates?

Add information about generating Priors / Estimates from observed data, summary statistics in the literature, ...

Add sampling options

Can we add some overview of other sampling options (also possibly to put in the R package)?
i.e. LAtin Hypercube (LHS) https://cran.r-project.org/web/packages/lhs/vignettes/lhs_basics.html , etc.
@PrajnaKARai and @johanneskopton

Evaluating participation of students in the class

A structured evaluation of the participation of the students concerning reading the provided materials and discussion in the class should be developed.

We could remove the multiple choice questions and introduce them in a Mentimeter, which the students can take at the beginning of the class, showing us how well they have understood the content ( 3 to 4 questions in 15 minutes max.). The concepts that have been misunderstood can then be discussed in detail.

Another activity that can be adopted to ensure students read the content posted for them is by providing one discussion question per group in the class. The student leading the discussion ( a different student every week) would present this to the class after 15 minutes.
The tutor could document the participation of the students.
This will form the base for the participation grade of the student.

"An error has occurred"

Getting a strange error on the web app again "An error has occurred"

The error point in the logs:

04/01 11:21:49.314 (GMT)
Error in value[3L] : parse error: trailing garbage
04/01 11:21:49.314 (GMT)
":{},"value":["0.10.1"]}]}]} {"type":"list","attributes":{},
04/01 11:21:49.314 (GMT)
(right here) ------^
04/01 11:21:49.314 (GMT)
Calls: local ... tryCatch -> tryCatchList -> tryCatchOne ->
04/01 11:21:49.314 (GMT)
Shiny R Markdown document exiting ...
04/01 11:21:49.314 (GMT)
Execution halted

The full logs:

rstudio-connect.234.0.1222.run_app.g6WZmicSktyEeDdt.log

Simplfy models in Seminar 9

The model is a bit too complex in the seminar model - hard for students to follow