2a:null 2b:["$","title",null,{"children":"RAMPVIS Idiom : OD Map | jason dykes | Observable"}] 30:T11c0,Flows are taken from the 2011 Census and acquired through UKDS - see [WICID](https://wicid.ukdataservice.ac.uk/). | Table number | **WU06BUK_la** | | --------------: | :---------------------------------------------------------- | | **Table title** | _Location of usual residence and place of work by industry_ | The 21 _WICID_ employment sectors are combined into 4 categories: | Name | Explanation | | ------------------: | :------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ | | **LikelyOperating** | _Industries in which workers are in key roles and so probably travelling under lockdown_ | | **OtherMiddle** | _Those employed in industries less likely to be working under lockdown, but who may begin working as lockdown loosens. Based on Scottish ordering of when people in particular roles are able to work._ | | **OfficeClosed** | _Those working in offices who are unlikely to be travelling during lockdown_ | | **OtherClosed** | _others in employment categories who are unlikely to be travelling during lockdown, including the retail, real estate, cultural and hospitality sectors._ | | **ALL** | All workers in all employment categories - the `normal' commuting flows according to the census. | In addition we use calculate a _signed CHI statistic_ to compare numbers of workers in each of these categories in relation to the overall figures. Negative numbers relate to lower than expected commuters in the category, give the overall proportion, in a way that accounts for magnitudes by under-emphasizing small signals. | Name | Explanation | | ------------------: | :------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ | | **CHI-LO** | _The weighted difference between the observed number of commuters in **Likely Operating** and those expected were workers who are **Likely Operating** distributed equally among **ALL** workers_
This lets us see the spatial effects of modelling lockdown commuting (model connectivity) versus regular commuting.| | **CHI-LOOM** | _The weighted difference between the observed number of commuters in **Likely Operating** and **Other Middle** sectors and those expected were workers in these sectors distributed equally among **ALL** workers_
This lets us see the spatial effects of modelling a less tight version of lockdown commuting (model connectivity) versus regular commuting, to see where differences persist as society opens up. | | **CHI-offC** | _The weighted difference between the observed number of commuters in **Office Closed** sectors and those expected were workers who are **Office Closed** sectors distributed equally among **ALL** workers_
This lets us see the spatial effects of modelling lockdown commuting (model connectivity) versus regular commuting by identifying places in which office closure has the greatest effect. | | **CHI-allC** | _The weighted difference between the observed number of commuters in **Office Closed** and **Other Closed** sectors and those expected were workers in the these sectors distributed equally among **ALL** workers_
This lets us see the spatial effects of modelling lockdown commuting (model connectivity) versus regular commuting by modelling by identifying places in which all forms of closure have strongest effect.| Outputs are appropriately protected so as not to identify individuals, households or organisations in line with UKDS conditions.
_Any cell count of lower than three has been replaced with zero_. 31:T477,**OD maps** are a rearrangement of the classic origin destination matrix in which rows and columns represent spatial entities and each cel is coloured according to some relationship between each pair. The rearrangement shows the spatial variation in relationships between each pair of spatial units. Flows, correlations, proximities are examples. In _OD matrices_, cells representing pairs of zones are positioned in rows according to origin and columns according to destination. In **OD maps**, the destination cells are arranged to produce small maps of the destinations of each origin, and coloured according to the numbers associated with each destination for a particular origin. Each of these destination maps for each origin is placed at a position representing the location of the origin in a larger map that is sized so that no occlusion occurs. **OD maps** work particularly well with regular (square) grids, and so transformation from uneven spatial units into a regular (square) grid, or _Grid Map_, may precede this process. Some _Grid Maps_ use gaps to retain aspects of the original spatial structure in the gridded output. 32:T4e2,We worked with modellers to select flows and derivations that allowed us to explore the nature of connectivity between locations at a number of scales and the effects of selecting particular employment sectors from the Census travel-to-work data to model these interactions. OD Maps were used as part of a design/redesign sequence that aimed to show geographic variation in flows in data that were highly skewed - as shown by the original OD matrices. This involved us using **GridMaps** to distort the geography used in our maps in a manner deemed acceptable to the task in hand. Here, geography is relative so distances and other spatial relationships are not preserved perfectly. The underlying geography and geographic relationships between spatial entities are sufficiently maintained for us to be able to see that the flow patterns vary spatially with some consistency however. We provided interactive OD maps - labels and quantities available as tooltips on mouseover - with log scaling. Log scaling is controversial as it is challenging to estimate quantities, which are very important in this context, but general structure is revealed. The question on which we focussed required us to consider both general structure and absolute numbers.33:T664, Our original maps of the NHS Board data showed commenting patterns to be highly local at that scale with most trips to work involving intra-board journeys. People tend to travel to work within area. Mixing between spatial units is very limited, though a distance effect was evident from the maps that we could not detect in the matrices. Flows tend to decrease with distance. That inter-board journeys tended to be local was not a great surprise - likely short trips across (arbitrary) boundaries in many case - but this was much more visible in the OD maps than in the OD matrix. This resulted in a change of scale of focus, to Local Authority. The intention here was to produce higher resolution OD maps, from a large OD matrix, and assess distributions in which flows between zones were more common. This allowed us to see some detail. The distance decay effect was evident, but not in the islands - where commuting to other LAs is sporadic and involves small flows and small numbers of people. Further North in Scotland, incoming flows tend to be skewed to the south, with more daily interaction between southern local LAs than those to the north. We saw big differences between flows of workers in ‘likelyOperating’ and ‘otherMiddle’ roles in the major cities - Dundee, Edinburgh, Glasgow & wider Lanarkshire. Much of this is within Local authority and so the value of the maps over the matrix is limited, but the spatial arrangement allowed us to see spatial clustering of large numbers showing national patterns of flows and that flows are relatively ‘local’ at this scale even when they occur across boundaries.34:T295e,## 7.0 Reflective Discussion OD maps are reported to have some advantages over the flow maps, as flow maps suffer from knowledge problems: * long distance flows (e.g. to/from Shetlands) having a disproportionate saliency (symbol saliency is not proportional to signal strength); * over-plotting makes (short, important) local flows hard to discriminate; * smaller regions of high population lead to a crowded view of flows and the occlusion makes it hard to see structure or detail; These were known reasons for implementing the OD maps, and they were shown to apply in our context, with the OD maps overcoming these problems effectively at the cost of losing geographic accuracy. The costs of this loss of geographic information are hard to establish with any precision, but the trade-off here seemed to result in some discussion and understanding that supported the user story and informed model inputs. Our collaborators strongly suggested that this position in the design space was a valuable way to add some spatial thinking to the consideration of the interaction matrix - a key input to the SNS model. Reactions from the modellers were positive:
"_Cool little-square OD maps! Much easier to see what’s going on than in matrices_" with some reassuring expected patterns identified and some question about appropriate levels of aggregation for modelling as we explored the spatial and scale-based structure of the connections. The maps worked well as a means of prompting discussion - the roles and spatial influence of Glasgow Airport became a point of discussion. The modellers reported that varying the interaction network might be less important at the coarser level when most movements are within region, but more important when looking at a finer level. The implications here were clearly described in our discussions: "_level at which we run implies **which aggregation** but ALSO **which slice** of commutes we need to use._ **<--- Important!**" There was some evidence of deep thinking about the spatial nature of phenomena that might have been prompted by the spatial representation:
“_what I'd really like to see is what happens after I account for spatial autocorrelation. It's not really news to know that nearby areas look similar_” So, in terms of the **user story** the OD Maps enabled us to see that the selection of commute ‘slices’ was more important as resolution increased. This is not surprising, but it was also evident that the effects were different in different areas, and where these effects varied. It’s a complex picture, but we did seem to find pictures that showed the complexity in ways that were interpretable. This allowed us to identify scales at which to consider importance of the problem in hand. However, despite the rich interpretable imagery provide by the OD Maps, this complexity made it very challenging to relate the data and the graphics to the final part of user story - decision-making. Moving from a detailed picture to action is difficult, and we moved from data to “we know a thing” much more smoothly than we were able to move from data to “_we know what to do_”. Turning complex visual answers into action is difficult. The visualization does improve understanding, but provide a general sense of “_how things look_” that is challenging to develop into an action. The evidence developed through the visualization results in a “_feeling_”. Is this adequate? It may well be effective decision-making, particularly where speed is of the essence! Many in the Simple Network Sim group had different roles, but these roles blurred in ADR (McCurdy et al., 2016)/ Design by Immersion (Hall et al., 2019) style. Despite the SEs not fitting the role described in the User Story very fully, we asked for feedback on the designs and their responses varied. One was highly skeptical and “_not sure if [OD maps] are immediately useful_”, but was also unsure as to whether the question in hand as important (with more of a focus on outputs). They described is as being “hard to see clusters when looking at the whole graph” and came up with many design ideas that pushed the focus elsewhere. In a way this emphasises the **need to design for particular tasks and users** and also that open design discussion documents such as those used here are an excellent means of engaging problem owners in discussions about the focus and nature of design. Our digital workspace, and the high priority work proved excellent at blurring distinctions between roles and problems aligning with notions presented in Hall et al. (2019) and Beecham et al. (2019). Another was more positive, and responded to thinking about the graphics and engaging with the data with:
“_I enjoyed doing this! Really interesting_”.
They offered simple accurate summaries:
“_Closing offices has a big effect on cities, but mostly on sort-distance commutes_”,
but emphasised the need for an overview: “_maybe I'm missing something important by filtering?_”.
They also provided an important reminder of the roles of labelling and legends in collaborative work upon which we were able to act rapidly:
“_the legends are a bit small!_” There was some focus on the visualization design rather than the data in our feedback, but this can be expected when learning about design and data are concurrent, and this felt healthy in terms of developing knowledge about visualization as well as engagement in the process and knowledge about the data. It also emphasized the blurry, personal and dynamic nature of a well-specified focus (or user story) and supported the need for interaction to personalise, adjust focus and back-track a careful transformation between views to ensure that changes of focus, data or visualization were not abrupt and thinking was not disrupted. One simple example in this case is that some data sets and questions required OD Maps and in other contexts DO maps were more revealing. In many cases, more complex visual transformations were necessary. Fluidity seemed important in our - and this may have been exacerbated or enabled by either or both of the digital workplace or the emergency context. The flow data were very peaky and so we used log scaling for some colouring. This is controversial and while log scaling helps us see structure, it is very challenging to interpret magnitudes, and both structure and quantity mattered in this user story. Interactive legends, switching between encodings and making stitches clear are very important as we deal with peaky and skewed data. We found the OD Maps to be very effective for showing regional clusters. While very radial, single centred distributions work very nicely with flow maps, where you have multi-scale phenomena with secondary clusters (e.g. Scottish MSOA travel to work data) and many local flows, OD/DO mapping is extremely informative. The ability to differentiate between incoming and outgoing flows (OD/DO) is also useful, this worked well in our efforts to address the question in hand. #### DIGITAL WORKPLACE We worked in GitHub, through email, with Zoom screenshare, videos dumped on Vimeo or elsewhere, and via Google forms / docs, all supported by threaded Zulip messaging. None of this is hugely important and other technologies might have worked equally well. But the fact that the workspace was entirely digital and that we were all locked to our screens did see to help with learning, focus and progress. Zulip, a searchable, threaded messaging system, meant that we could explain and link and comment freely and easily. The ability to share data and development on GitHub, and style pages to explain and generate discussion and feedback with dynamic documents (see Beecham et al., 2019) proved highly effective, enabling rich, rapid, considered responses to be generated asynchronously. This was very effective. #### EMERGENCY VISUALIZATION The designs were developed as contributions to a disparate group of volunteers who were previously unknown to one another. Roles were relatively ill-defined and volunteers had varied considerably in terms of their skills, focus, levels of experience with data, objectives, motivations, interest, time. All were working in parallel, with volunteers arriving and departing mid-project and all participants dealing with the uncertainty, pressure, worries (finance, food, lines), sickness and caring responsibilities associated with an unprecedented pandemic and multiple people This meant that we needed alternative views and perspectives. It would be nice to think that visualization helped with this, providing a gateway into the problems and perhaps even common boundary objects that showed progress, shone light onto the problem and brought people together. It would be nice to think that the graphics introduced people to the problem, status and priorities as they joined the team. I am not certain that this is true, we don’t have the evidence to make this claim. But, one recommendation would be that an annotated visualization noticeboard of “_what we know_” and “_what’s new_” might have been a really good way to get people up to speed and structure progress. We didn’t work as formally as that. It was also very clear that any graphics needed user control, to refine them to particular needs, but also rapid introduction and learning to that team members could use these interactive graphics very quickly. A suite of possible ‘views’ and interactions seems important, but this must come with strong support. Giving people everything meets all needs, but also makes systems less usable. OD Maps and perhaps other novel views need guided introduction - again , a DExD style setting (Beecham et al. 2019) may be useful. That was very achievable here given workspace (virtual) and tech (Web/JavaScript). Having ‘wizard of Oz’ interfaces in which visualisation designers provided the knowledge also worked well in the digital workspace. 35:T42b,## 7.3 RECOMMENDATIONS & OPPORTUNITIES - Use OD maps to show the spatial structure of irregular flows - with multiple sources and destinations, and varying flow quantities and lengths. - OD maps work well for tens or hundreds of spatial units. - Grid maps work well as OD map layouts - Make flow maps a viable and accessible alternative - Ensure that appropriate options for scaling, filtering, design and interaction are available & accessible (e.g. log / regular scaling) - Use ‘Wizard of Oz’ to give rapid access to alternative representations in a fluid manner - this is good use of visualization expertise - Use interactive notebooks with structured explanations and direct requests for responses to develop shared understanding of design, data and solution development - Use a searchable threaded means of asynchronous discussion to log reactions, offer explanations and develop data and design discourse. - Develop a “_what we know_” annotated visualization noticeboard with buy-in from project lead to log progress and get volunteers up to speed. 36:[] 2e:{"id":"7f5c259702670aad","slug":"rampvis_idiom_odmap","title":"RAMPVIS Idiom : OD Map ","thumbnail":"bdc7a3af7e5de3f6f3f0b7141331ef7e3d917556cccfd40d3619deda7fc3aac1","access_level":"unlisted","create_time":"$D2021-07-30T10:58:04.057Z","update_time":"$D2021-08-08T10:55:38.383Z","view_time":null,"first_public_time":null,"has_unpublished_edits":null,"fork":{"id":"5f69cc3361f8f87f","slug":"rampvis_idiom_odmap","title":"RAMPVIS : The OD Map","owner":{"id":"9461f58b209074d5","login":"jwolondon","name":"Jo Wood"}},"stars":0,"starred":false,"owner":{"type":"team","id":"262e0f3061f34839","login":"jsndyks","name":"jason dykes","avatar_url":"https://avatars-next.observableusercontent.com/ce8c740e3f4f161e440b04f5d1f16368eb63cba559ce4c7bdec538f4316764d8","tier":"starter_2024"},"version":440,"publish_version":440,"role":null,"authored":false,"authors":[{"type":"individual","id":"955e37df990d4e95","login":"jsndyks","name":"jason dykes","avatar_url":"https://avatars-next.observableusercontent.com/ce8c740e3f4f161e440b04f5d1f16368eb63cba559ce4c7bdec538f4316764d8","tier":null}],"members":[{"type":"individual","id":"955e37df990d4e95","login":"jsndyks","name":"jason dykes","avatar_url":"https://avatars-next.observableusercontent.com/ce8c740e3f4f161e440b04f5d1f16368eb63cba559ce4c7bdec538f4316764d8","tier":null,"role":"editor"}],"comments":[],"custom_thumbnail":false,"trashed":false,"lapsed":null,"subscribed":null,"forks":1,"collection":null,"type":"notebook","body":{"titling":"auto","title":"RAMPVIS Idiom : OD Map ","cells":[{"id":0,"value":"# RAMPVIS Idiom : OD Map
\nThis is the **actual one**, rampvis_idiom_odmap, to be used as data in, and supplementary materials for, the _TransPhilA_ paper
\nIt's not the original **example one**, which is here\n
https://observablehq.com/@jsndyks/rampvisexample_odmap?collection=@jsndyks/rampvis/2","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":4,"value":"---\n## 1. TASK & USERS - User Story\n\n * As **an epidemiological model builder**\n * I would like to\n * **see the spatial patterns of commuting among different employment sectors**\n * **understand how these differences vary spatially and the scales at which they operate**\n * so that I can \n * **select appropriate spatial interaction matrices for my simulation models that relate to different lockdown scenarios**","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":128,"value":"---\n\n## 2. DATA\n\n#### Spatial Data - for N units\n\nGrid positions and populations for each of the spatial units:\n * https://raw.githubusercontent.com/gicentre/scrc/main/data/grid/scotLAsGrid.csv\n\n#### Relationship Data - for < N*N relationships\n\nPairs of spatial unit identifiers followed by any number of numeric values representing relationships between them:\n * https://raw.githubusercontent.com/gicentre/scrc/main/data/flows/scotLAsFlows.csv\n\nHere we have columns for _all commuters_, then three different groups of commuters representing alternative lockdown scenarios as described below. The final three columns compute some differences between scenarios. Data are derived from 2011 Census of Population. ","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":136,"value":"spatialData = await d3\n .csv(\n \"https://raw.githubusercontent.com/gicentre/scrc/main/data/grid/scotLAsGrid.csv\"\n )\n .then((objs) => objs.map((o) => Object.values(o).map(String)))","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":141,"value":"odData = await d3\n .csv(\n \"https://raw.githubusercontent.com/gicentre/scrc/main/data/flows/scotLAsFlows.csv\"\n )\n .then((objs) => objs.map((o) => Object.values(o).map(String)))","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":305,"value":"$30","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":7,"value":"---\n\n## 3. NAME - and description","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":43,"value":"$31","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":10,"value":"---\n## 4. ORIGIN & CONTEXT\n\n**OD Maps** were introduced by _[Wood et al (2010)](https://doi.org/10.1179/000870410X12658023467367)_ as a likely intuitive and effective spatial means of representing the geography of geographic variation.\n\nSubsequently, _[Yang et al (2016)](https://doi.org/10.1109/TVCG.2016.2598885)_ found them to be as effective, or more effective than alternatives for complex map tasks in their study, in which they were deemed '_most readable_' by participants in their experiments.","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":13,"value":"___\n\n## 5. EXAMPLES\n\n**City Intelligence** use an OD Map to show how commuter flows vary in the boroughs of London :\n * [_London Commuter Flows_](https://maps.london.gov.uk/commuter-flows/)\n\nTableau Zen Guru **Robert Radburn** uses OD Maps to show inter-state migration in Africa :\n * [_Where migrants came from and where migrants went in Africa_](https://public.tableau.com/app/profile/robradburn/viz/AfricanODMigration/Wherewecamefromandwherewewent)","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":53,"value":"---\n## 6. DESCRIPTION & USE","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":336,"value":"$32","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":340,"value":"**FIGURE** -- \n_See **Outgoing Flows** on the [Visualizing travel-to-work patterns](https://gicentre.github.io/scrc/docs/flow/allScotland.html) page_
\n**But we need this for NHS Boards!**","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":348,"value":"$33","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":353,"value":"**FIGURE** -- \n_See **Outgoing Flows** and **IncomingFlows** under **4. OD Maps** on the [Visualizing travel-to-work patterns](https://gicentre.github.io/scrc/docs/flow/allScotland.html) page_
\n**It would be good to show the OD Map, DO Map and the OD Matrix side-by side here**","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":394,"value":"We can see nuanced patterns too - for example when we map the differences between numbers in the _Likely Operating_ sectors and what we would expect if these flows were in proportion to _all_ commutes. Patterns tend to be local, are strongest in the urban areas, but vary. The East of Scotland tends to have fewer commuters _Likely Operating_ than we would expect given the overall numbers, with effects strongest in and around Aberdeen and Edinburgh. ","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":325,"value":"fig1 = fetchImage(\n \"https://gicentre.github.io/scrc/docs/simpleNetworkSim-VIS/img/snsVIS.v5.pic.00397783.png\"\n)","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":360,"value":"We needed more details but OD maps are unlikely to be interpretable with high numbers of spatial units. So we selected areas of interest and increased resolution - effectively a filter and split operation to focus in on the detailed flows at more local level in places of interest.\nOur focus here was Glasgow and Strathclyde, and we used MSOAs as a detailed but manageable level of resolution.\n\nMany of the big flows between MSOAs are quite local, but most MSOAs send people to a relatively wide range of destinations. At this scale, the inter-unit flows are much more significant, with important implications for the modelling: varying interaction modelling will have effect at this scale.","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":370,"value":"**FIGURE** -- \n_See **https://gicentre.github.io/scrc/docs/flow/glasgow.html** on the [Glasgow / Strathclyde travel-to-work patterns](https://gicentre.github.io/scrc/docs/flow/glasgow.html) page_
\n**It would be good to get OD and DO maps side by side here for ALL!**","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":380,"value":"In terms of the spatial structure of workers in the _Likely Operating_ sectors, the OD Maps of outgoing flows (destinations for each origin) show that in almost all cases (all sure MSOAs) people travel widely across the region. In terms of incoming flows, a DO Map, of origins for each destination produced mini maps of real focus, with varying degrees of local spatial structure. These reveal focussed and important centres of employment that vary in the degree to which they attract local workers from local or more distant areas: they include Battlefield, Paisley, Govan, Springburn, Kelviside, each of which attracts many workers from a wide catchment.\n","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":375,"value":"**FIGURE** -- \n_See **https://gicentre.github.io/scrc/docs/flow/glasgow.html** on the [Glasgow / Strathclyde travel-to-work patterns](https://gicentre.github.io/scrc/docs/flow/glasgow.html) page_
\n**It would be good to get OD and DO maps side by side here for Likely Operating!**","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":170,"value":"\nDifferent types of job have very different spatial structures at this scale, with local centres much more likely to employ those in ‘likelyOperating’ roles and the city centre being a focus for ‘officeClosed’ jobs. Note the very different kinds of work in adjacent City centre MSOAs shown below.","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":332,"value":"fig = fetchImage(\n \"https://gicentre.github.io/scrc/docs/simpleNetworkSim-VIS/img/snsVIS.v5.pic.00054812.png\"\n)","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":386,"value":"Using OD and DO maps to show travel to work patterns by employment sector strongly suggest that at MSOA level differences in daily travel between employment sectors are pronounced and inter-zone mixing is extensive meaning that the input network will be sensitive to these decisions at this scale.\n\n---","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":225,"value":"### Unfinished / To Do!","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":401,"value":"**JO** offered to develop the OD Maps in this idiom! ","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":56,"value":"{\n reset;\n return exampleODmatrix();\n}","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":75,"value":"viewof reset = Inputs.button(\"Reset\")","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":73,"value":"viewof fCol = Inputs.range([2, 7], {\n value: 2,\n step: 1,\n label: \"Flow Column\"\n})","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":84,"value":"viewof layout = Inputs.range([1, 3], {\n value: 1,\n step: 1,\n label: \"Alternative Grid Layouts\"\n})","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":90,"value":"---\n## 7. REFLECTION","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":408,"value":"$34","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":414,"value":"## 7.1 WHAT WORKED\n - OD Maps worked well in addressing the question in hand and provided rich outputs that showed the data fully and allowed us to see the effects of employment sector selection on the interaction model, and how this varied with scale and geography. \n - Grid maps are useful frames for OD maps\n - The digital workspace meant that visualisation designs could be tweaked, chawed, run and interacted with on the fly. This weather very effective.\n - The insightful graphics worked in terms of developing mutual trust and confidence in capability. Showing something new is motivating, particularly if it is meaningful.","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":422,"value":"## 7.2 WHAT DIDN’T WORK\n\n- OD Maps were not embraced by all members to the team, but then did result in useful discussion around needs, priorities and visualization possibilities in all cases.\n- Developing specific actions from rich visual descriptions of data is a challenge, The move from seeing and understanding, to doing in light of complexity is challenging.\n","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":426,"value":"$35","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":434,"value":"## 7.4 WHO WAS INVOLVED\n\n - **Jess Enright** - epidemiologist, coordinator of Simple Network Sim model\n - **Fedor Gorokhovik** - software engineer, Simple Network Sim\n - **Robert (Bob) Turner** - lead software engineer, Simple Network Sim\n - **Jason Dykes** - visualization support & design\n - **Jo Wood** - visualization support & design\n ","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":16,"value":"---\n\n## 8. REFERENCES","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":27,"value":"**Wood, J., Dykes, J., & Slingsby, A. (2010).**
\n_Visualisation of origins, destinations and flows with OD maps._
\nThe Cartographic Journal, 47(2), 117-129.
\nhttps://doi.org/10.1179/000870410X12658023467367","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":270,"value":"**Yang, Y., Dwyer, T., Goodwin, S., & Marriott, K. (2016).**
\n_Many-to-many geographically-embedded flow visualisation: An evaluation._
\nIEEE Transactions on Visualization and Computer Graphics, 23(1), 411-420.\nhttps://doi.org/10.1109/TVCG.2016.2598885","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":57,"value":"---\n### Imports & Functions","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":66,"value":"import { exampleODmatrix, animTimer } from \"@jsndyks/rampvis_exampleodmatrix\"","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":77,"value":"timeToComplete = {\n return 25000;\n}","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":79,"value":"animProp = {\n reset;\n return animTimerNew(timeToComplete);\n}","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":298,"value":"function* animTimerNew(ttC, timeDelay = 5) {\n const timeBeg = Date.now(); // Poll the current time in milliseconds.\n const timeEnd = timeBeg + ttC;\n let t = 0;\n let p = 0;\n yield t; // Yield an immediate result at start\n while (t <= 1) {\n t = (Date.now() - timeBeg) / (timeEnd - timeBeg);\n // -- Work out bound proportion!\n // p = Math.round((1000 * (Math.sin(Math.PI * 2 * t) + 1)) / 2) / 1000;\n p = Math.round(1000 * Math.sin(Math.PI * 2 * t)) / 1000;\n // Yield subsequent results after the given delay.\n yield Promises.delay(timeDelay, Math.max(Math.min(1, p), 0));\n }\n}","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":317,"value":"import { fetchImage } from \"@mbostock/cross-origin-images\"","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":202,"value":"---\n\n### Styling","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":207,"value":"","pinned":true,"hidden":false,"mode":"html","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"}],"files":"$W36","stdlib":"1"}} 2f:"$2e:body"