observe and understand?

Very clear!

Can you link directly to the data? (should be open on SCRC site). Can you give info on what acronyms mean? cumVaccination, etc. Be reader friendly!

This is core "information visualization" as originally defined - can you use the term? "spaces" repeats. "more easily" is contentious as you lose detail ... can you address this tension in the description - we need sweet spots! Can you make this clear?

Nice description. Feels close the the 'algorithms' workbook though. This is a series of 2D projected data spaces right? https://static.observableusercontent.com/files/3567fe6a605656acdc14eaefb7cd335310c6dc9a069596a0437b6a4d487787026f25d79f4f75cd9c4bbce902ec396f4a35537cb03b08756cfa436cec22d14601 Projection at two levels - scatter plots, within grid layout?

Added a link to that other idiom, they are pretty distinct in terms of their emphasis/application but a link worked here just fine to build a bridge to that.

Can you make this more of a narrative - references come later. See https://observablehq.com/@jsndyks/rampvis_idiom_odmap Check this observable for ideas on section 3 too.

Addressed in the previous revision.

What did you achieve in relation to the User Story? Figures should show what you found - not just what you did!

Improved this significantly now.

Nice . Can you split into blocks? "relatively higher" - relative to what, higher than what? I'd just use "high".

Can you reorder the matrix to show something more useful for interpretation than the structure of the Alphabet? By population perhaps?

Matrix - is there similarity / redundancy, do we want to show correlation with self, do we want to comment on this?

Interesting - I am a bit skeptical about the "lag-corrected-data-space" being a "derived data space" as it's a transformation in a space (-T) not a new space ... thoughts?

Of course, I would love to see an OD map of -T for all regions, (small maps) for all regions (large map). Do you have the data? Tlag for all pairs of regions?

I'd be tempted to cut "Developing a time-lag-corrected data space for effective time-series comparison" - down to you! I think it blurs the focus and message.

"Dynamic Time Warping" is mentioned once. Can you explain it? Sounds very Star Trek. We need to know what it is!

Lag graph - show the lag! You have a timeline! Put it on the axis with a line representing lag offset to compare data spaces!

I don't think we would like to take down the whole time-lag analysis. I think it adds diversity and an interesting twist that broadens the scope of what is meant by data spaces (one can even think of simple algebraic operations (timeSeriesA - timeSeriesB to find differences) as a way of deriving a new space (e.g, difference space -- where each data element represents difference). I think the time-lag transformation/derivation provides an example of this nicely.

Added an explanation of Dynamic Time Warping and a reference - it does sound very star treky indeed..

Also split this into blocks.

Also added an annotated figure with the lag indicated, good suggestion!

Franck also did a reordering of the heatmap, which worked nicely.

It's quite hard to get a sense of what this means without examples or figures. What kinds of structures? How did they help with the user story?

Nice!

That's pretty problematic and pretty predictable! So it sounds as though it was NOT easy as expected (see section 3) and the User Story was not achieved: "so that I can observe the differences within the progression of the pandemic". > We can make a new thing... nobody understands what it is. So what do we learn? How can we do better? What do you recommend? I'd love some deeper thinking here. ↓😉 You have valuable experience. Please chat with each other and modelling colleagues to work out what this means!

A lot of new thinking here which was not possible to communicate earlier due to limited time. Hopefully this should give more to bit into now.

Great: "Not having regular access to epidemiologists" is part of the "idea generation lab" that we need to factor in to our processes! How? We should chat!

Why underwhelming? For whom? Not underwhelming if it shows something useful? (confirmation, ideation)?

Very good points, addressed these now a bit in the discussion here, but would be good to expand later on as well.

Delete

Delete the instructions!

And unpin all markdown!

Great reflection. Love the notion of Analytical Design Narrative, which ties in well with the DExD and LitVis stuff we have been writing.

Great - I like : "Knowing what is possible computationally" - this is really important knowledge that is missing (I think) from DSM where "learn" is mainly about design ... but I need to check this!

Great work guys. Update in BibTex and make sure names are in the spreadsheet!

Very nice - great reflection! Please see comments and add examples. I might consider removing “Developing a time-lag-corrected data space for effective time-series comparison”. Down to you, but see comments! Apologies for design suggestions - I cannot help it when I see scope for encoding useful additional information with few draw-backs!

All resolved, thanks for the suggestions. We are exploring whether there is a possibility to make a change to the designs of some of the visualisations (such as ordering) but for now that might need to be put on hold.