2d:null 2e:["$","title",null,{"children":"Voyager 2 Summary | Edmund Wong | Observable"}] 33:T495,md`## Design Considerations The research paper also talks about several design considerations. This part of the paper delves into a lot of the thought process behind a lot of the decisions made. I personally find these design considerations really interesting. One of the design considerations is to prioritize showing different variants of the data as opposed to just different designs. Different users should be able to use this in the way that they want to so Voyager 2 allows interactivity through a "Tableau-style" drag and drop interface. In addition, this is supplemented by related views that you can bookmark or add as the specified view. Voyager also supports undo and redo for these changes. Voyager aims to create expressive and effective visual encodings. This is to try to reduce misleading charts by following perceptual design principles. In order to prevent showing every possible chart, Voyager 2 includes interactions for sorting, scale tranforms, and axis transposition. These are supported in the top toolbar section of each visualization. To prevent redundant charts, Voyager 2 suggests related views with data variation over design variation. `34:[{"id":"7f7a15ea3d9a6718bb1e7effc39a30cd764b4bc9bd169a463abb3e796226adbe751c2bdd9d8f5e91424b590a1da74189c67c2e528ba336e7bea2c89584211502","href":"https://static.observableusercontent.com/files/7f7a15ea3d9a6718bb1e7effc39a30cd764b4bc9bd169a463abb3e796226adbe751c2bdd9d8f5e91424b590a1da74189c67c2e528ba336e7bea2c89584211502","name":"univariate.png","type":"image/png","lastModified":1585767699828.305,"size":105346},{"id":"02a618dc58fac953993dd3c351bdd0458d140e5e39d45dc929e3ce96d32dfe846a44086aa302d7df6c374233e5ee2493e3020eacda73e0342b781cf9f780b1db","href":"https://static.observableusercontent.com/files/02a618dc58fac953993dd3c351bdd0458d140e5e39d45dc929e3ce96d32dfe846a44086aa302d7df6c374233e5ee2493e3020eacda73e0342b781cf9f780b1db","name":"Screenshot (31)@1.png","type":"image/png","lastModified":1585767775703.344,"size":403740},{"id":"75b62b58a2c227280f84d59a9beaaaa0ac25a7967660ac93dd38c4e02814bf3c1e87804ce12ce57d41fc235d9967c0a7a6540a050a115d12182fccde28aea2c7","href":"https://static.observableusercontent.com/files/75b62b58a2c227280f84d59a9beaaaa0ac25a7967660ac93dd38c4e02814bf3c1e87804ce12ce57d41fc235d9967c0a7a6540a050a115d12182fccde28aea2c7","name":"quantitative.png","type":"image/png","lastModified":1585769876421.345,"size":186039},{"id":"3ac362688f8bbef934f7afdf54da22d17319d42b777a22ddd4f3d656374f5961ff31bb7f4f1c0aa785670977f69eebcf2aa85f5b09f557d07f168e8eebfd64a0","href":"https://static.observableusercontent.com/files/3ac362688f8bbef934f7afdf54da22d17319d42b777a22ddd4f3d656374f5961ff31bb7f4f1c0aa785670977f69eebcf2aa85f5b09f557d07f168e8eebfd64a0","name":"wildcard.png","type":"image/png","lastModified":1585796485623.628,"size":288204},{"id":"e2e0ae00995dd90fb9cad805c3f93b45ece1e916870fc3d1ea8d77965fee65973ad2794a883cb91540cf832843a61f03cf02a16e75c4a4562eee72973888d3fe","href":"https://static.observableusercontent.com/files/e2e0ae00995dd90fb9cad805c3f93b45ece1e916870fc3d1ea8d77965fee65973ad2794a883cb91540cf832843a61f03cf02a16e75c4a4562eee72973888d3fe","name":"specified.png","type":"image/png","lastModified":1585768543505.307,"size":70026},{"id":"ee56991d17d689f597b4dafa1df591ba99095231b1cfc1a83bb4c7cbc21b3e43241d5e2419f3f5cdd435b0eb1027c4d4fff9d1e61318c0f4b10769d18ff895bc","href":"https://static.observableusercontent.com/files/ee56991d17d689f597b4dafa1df591ba99095231b1cfc1a83bb4c7cbc21b3e43241d5e2419f3f5cdd435b0eb1027c4d4fff9d1e61318c0f4b10769d18ff895bc","name":"categorical.png","type":"image/png","lastModified":1585768543504.267,"size":91906},{"id":"22fbee0af5be68c1103815205f3e7294f3cdc6695ff1aff9bcf7f1eac7bb2acbd6375695a4c45554af690c9160e0b2c20278c794ab1f8b0fd5ae07a2982bb1f3","href":"https://static.observableusercontent.com/files/22fbee0af5be68c1103815205f3e7294f3cdc6695ff1aff9bcf7f1eac7bb2acbd6375695a4c45554af690c9160e0b2c20278c794ab1f8b0fd5ae07a2982bb1f3","name":"summaries.png","type":"image/png","lastModified":1585768543505.393,"size":127943}] 31:{"id":"f3d90dd1a1ef1163","slug":"voyager-2-summary","title":"Voyager 2 Summary","thumbnail":"791d744255ad0018c97dfef28429f5867b312db6194b473265c4b26aedb6869b","access_level":"public","create_time":"$D2020-04-01T18:19:37.724Z","update_time":"$D2020-04-02T19:15:49.836Z","view_time":null,"first_public_time":"$D2020-04-01T18:20:20.625Z","has_unpublished_edits":null,"fork":false,"stars":0,"starred":false,"owner":{"type":"team","id":"25c713a14090b99f","login":"ewwong2","name":"Edmund Wong","avatar_url":"https://avatars-next.observableusercontent.com/3470f0a32fd4854755cea63e1c11dbc6884d794b03ccf2d26560d49a4985dc3c","tier":"starter_2024"},"version":168,"publish_version":168,"role":null,"authored":false,"authors":[{"type":"individual","id":"a6238ad8f8dcb07f","login":"ewwong2","name":"Edmund Wong","avatar_url":"https://avatars-next.observableusercontent.com/3470f0a32fd4854755cea63e1c11dbc6884d794b03ccf2d26560d49a4985dc3c","tier":null}],"members":[{"type":"individual","id":"a6238ad8f8dcb07f","login":"ewwong2","name":"Edmund Wong","avatar_url":"https://avatars-next.observableusercontent.com/3470f0a32fd4854755cea63e1c11dbc6884d794b03ccf2d26560d49a4985dc3c","tier":null,"role":"editor"}],"comments":[],"custom_thumbnail":false,"trashed":false,"lapsed":null,"subscribed":null,"forks":1,"collection":null,"type":"notebook","body":{"title":"Voyager 2 Summary","cells":[{"id":0,"value":"md`# Voyager 2 Summary\nby Edmund Wong\n`","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":4,"value":"md`\n## Paper Source\n\n[Voyager 2: Augmenting Visual Analysis with Partial View Specifications](https://idl.cs.washington.edu/files/2017-Voyager2-CHI.pdf) \nKanit Wongsuphasawat, Zening Qu, Dominik Moritz, Riley Chang, Felix Ouk, Anushka Anand, Jock Mackinlay, Bill Howe, and Jeffrey Heer\n\nPublished in: [ACM Human Factors in Computing Systems (CHI), 2017](https://dl.acm.org/doi/abs/10.1145/3025453.3025768)\n\n[Demo](https://vega.github.io/voyager2/) | [Source](https://github.com/vega/voyager)\n`","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":8,"value":"md`## Summary\n\nVoyager 2 is a visualization tool for data analysts that allows for multiple moes of interaction. It allows for focused analysis and targeted question answering with drag and drop featues in a \"Tableau-styled\" way that allows you to customize your view of the data. In addition, it also allows for open-ended exploration. This exploration is encouraged through related views that are generated based on the focus view.`","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":17,"value":"md`## Layout`","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":13,"value":"html`\n

\nThis screenshot is of the demo using the Iris dataset. This includes multiple layouts such as the data with fields, encoding, specified view, and related views.\n

\n`","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":67,"value":"md`### Univariate Summaries`","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":44,"value":"html`\n

\nThe related views section contains a lot of different views. One view is the univariate summaries. In this case, we can see a couple different encoding for each field in the dataset.\n

\nThe first chart is a horizontal bar chart encoded by species and count of species. The other charts are vertical bar charts that bin the quantitative data and show the count for each of the remaining fields.\n

\n`","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":70,"value":"md`### Specified View`","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":46,"value":"html`\n

\nIn the first layout picture that was shown, the x was petalLength and y was petalWidth. In this screenshot, we can see that our specified view is encoded as such.\n

\n`","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":73,"value":"md`### Summaries`","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":47,"value":"html`\n

\nOnce we select our encodings, we get a bunch of related views that change based on the encodings that have been chosen. In this image, Voyager 2 is showing different ways to view the same data in a bunch of different summarized ways. This includes using mean and bins. \n

\n`","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":76,"value":"md`### Categorical Field`","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":49,"value":"html`\n

\nIn the categorical related view, we can see that Voyager 2 has added the species as the categorical color field. In this case, they are using a scatterplot encoded by our original encodings in our specified view.\n

\n`","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":79,"value":"md`### Quantitative Field`","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":65,"value":"html`\n

\nAnother related view is the quantitative field view. This adds the quantitative fields as a size to the specified view. In this case, sepalLength and sepalWidth have been added to respective charts.\n

\n`","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":89,"value":"md`## Wildcards`","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":92,"value":"md`One particularly interesting part of the paper is the use of wildcards. This allows multiple charts to be shown based on the wildcard(s).\n`","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":97,"value":"html`

`","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":113,"value":"md`In this example, two quantitative wildcards were selected in the data section. Note that the Iris dataset has 4 quantitative fields. This results in 6 graphs (4 choose 2). The charts are scatter charts for each point in the dataset. The axes are encoded by each combination of the quantitative data fields.`","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":117,"value":"md`This wild card functionality is supported by [CompassQl](https://github.com/vega/compassql). This results in a spec that looks like the following:\n`","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":139,"value":"({\n \"spec\": {\n \"data\": {\"url\": \"data/iris.json\",\"formatType\": \"json\"},\n \"mark\": \"?\",\n \"transform\": {\"filterInvalid\": true},\n \"encodings\": [\n {\"channel\": \"x\",\"field\": {\"name\": \"fx\"},\"type\": \"quantitative\"},\n {\"channel\": \"y\",\"field\": {\"name\": \"fy\"},\"type\": \"quantitative\"}\n ],\n \"config\": {\n \"overlay\": {\"line\": true},\n \"scale\": {\"useRawDomain\": true},\n \"facet\": {\"cell\": {\"width\": 150,\"height\": 150}}\n }\n },\n \"groupBy\": \"field\",\n \"orderBy\": [\"fieldOrder\",\"aggregationQuality\",\"effectiveness\"],\n \"chooseBy\": [\"aggregationQuality\",\"effectiveness\"],\n \"config\": {\"omitTableWithOcclusion\": false,\"autoAddCount\": false}\n})","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":144,"value":"md`Note the use of \"?\" for wildcards. This can be found in the debug section of the website. The debug section also shows the Vega specification used to create the chart.`","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":149,"value":"$33","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":156,"value":"md`## Conclusion\n\nClearly from the paper, Voyager 2 has the makings of a very powerful visualization and data analytics tool. This tool supports everything from manual specification to suggestions with related views. This included being able to create quick prototypes or targeted question answering. The source of the paper as well as the demo of the tool can be found at the top of the notebook. I would highly suggest testing out the tool and explore Voyager 2 yourself.\n`","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"}],"files":"$W34","stdlib":"1","titling":"auto"}} 32:"$31:body"