1a Data Types.pdf

Data Types 1 From “Visualization Analysis & Design” T. Munzner, CRC Press, 2015 (Chapter 2) 2  Five different data types item: an object link: relationship bet...

Data Types 1 From “Visualization Analysis & Design” T. Munzner, CRC Press, 2015 (Chapter 2) 2  Five different data types item: an object link: relationship between items attribute: property of an item position: a location in 2D or 3D space grid: regular sampling of continuous data grid: more of an approach to collecting and storing data than a data type itself 3 Running example 4  “Running” example Hill running in Scotland Runners take part in races Races are held annually Scottish Hill Racing: https://www.facebook.com/scottishhillracing/ 5  Five different data types item: a runner link: two runners train together (“run-buddies”) attribute: a runner belongs to a club position: the start point of a race grid: a runner’s heartbeat sampled every 30s 6  Four different data set types A data set type is a method for collecting data together – table: rows and columns (2D or multidimensional) – networks and trees: relationships between items – fields: continuous data (conceptually there are an infinite number of measurements you could take, so sampling and/or extrapolation are necessary) – geometry: spatial data 7  Data set type: table Table: rows and columns (2D or multidimensional) John Dunne m M50 Springly Sara Ahmed f F60 Ludders Mei Chan f F40 Bowlerside Charles Ndlovu m M35 Ludders 8  Data set type: table Table: rows and columns (2D or multidimensional) y Springl y Springl y y Springl Springl y John Dunne m M50 Springly Springl Ludder s Panton Sara Ahmed f F30 Ludders Panton Panton Panton Mei Chan f F40 Bowlerside Ludder s s Charles Ndlovu m M30 Ludders Ludder Modelling changes over time — e.g. adding in the ‘key’ or dimension of years, to represent changes in clubs (and age/gender categories) — makes the table multidimensional 9 Data set type: networks and trees Networks and trees: relationships between objects Links show run-buddies Run-buddies are static pairs, but those pairs can group together in race events 10  Data set type: fields Fields: continuous data. Conceptually there is an infinite number of measurements you could take, so sampling and extrapolation are necessary every 10s 90bpm 10s 90bpm 100bpm 25s 102bpm 105bpm 26s 103bpm 106bpm 40s 106bpm 110bpm 52s 112bpm 115bpm 58s 114bpm 135bpm 73s 137bpm 140bpm 80s 140bpm A runner’s heartbeat. If you do not know the frequency of the heartbeat, e.g. every 10 seconds (as in the left example), then it has to be an attribute of the data item. Then you might use a table with two columns to represent the measurements as well as the times they were taken, as in the central example here. The example on the bottom right is from Munzner’s book, and I think it's to show a polar coordinate style sampling of a continuous region of space, i.e. positions are based on four distances from a central point, and an arc of 16 angles around that central point, so that the region has 64 samples... and so 64 cells. Each cell has 5 values measured or calculated for it, and we can see one in the bottom of the figure. One value in that cell is highlighted. It's unfortunate that the same colour used to highlight the value in the cell is also used to highlight the cell in the field. I would guess that Munzner is using this arc shape just to show that fields don't always have to be based on rectangular shapes or orthogonal axes. It's perhaps also unfortunate that the book doesn't explain this. 11  Data set type: geometry Geometry: spatial data Location of the annual Hill Running Races in Scotland, by start point Note that if the name of a location might not be enough to be spatial data, partly as the name might be ambiguous. Also, names such as locations of hill races would have to be mapped on to a coordinate system, such as latitude and longitude (and perhaps altitude too), before we could call it spatial data. 12  Data Availability Data is available at the same time, or collected as as dynamic stream Not the same as ‘data with a time dimension’ ‘Online’ or ‘Offline’ Average finish time for the Two Average finish time for the Two Breweries race in 2018 Breweries race over all time The assumption is that usually the data would all be available at the same time; streaming gives its own challenges 13  Attributes Attribute types and Ordering direction 14  Attribute Types club: Springly, Ludders, Bolderside, Sharpford. race difficulty: very difficult difficult managable by most runners easy very easy finishers’ time: 1h40, 1hr42, 1hr53, 1hr54, 1hr58… race date: 10th April, 15th Apr, 3rd May… Categorical – also called ‘nominal’ – no implicit ordering, can only say whether things are the same or different. Can impose an ordering (e.g. alphabetic, the number of members in a club), but this is imposed, not intrinsic Ordinal has an intrinsic order, but the distances between items are not determined… so they cannot be added or subtracted Quantitative involves a metric space, and so they can be added, subtracted (and – usually but not always – divided). 15  Ordering direction Finisher’s time for a race: 1h40, 1h42, 1h53, 1h54, 1h58… elevation: 100m below 50m below 50m above race date: 100m above 10th April 15th Apr 3rd May… …11th December 8th April 10th April Race date is cyclic data…. although, of course, if you included the year in your Race Dates, it will be sequential data 16  Running example: Two Breweries Hill Race (TBHR) Year Position Bib number Name Club Age category Finishing time 17  Running example: Two Breweries Hill Race (TBHR) Year, Position, Bib number, Name, Club, Age category, Finish time 1984 1 69 J Maitland Aberdeen ACC MOPEN 2:44:36 1984 2 53 B Brinfle Horwich RMI M50 2:50:36 1984 3 64 ARJ Curtis Livingston & D W50 2:52:34 1984 4 24 S Moore Horwich RMI M40 2:53:01 1984 5 65 AW Spenceley Carnethy HR WOPEN 2:56:55 1984 6 77 M Lindsay Carnethy HR MOPEN 2:58:42 I will leave it up to you to determine the data categories for each column 18  Running example: Two Breweries Hill Race (TBHR) Year, Position, Bib number, Name, Club, Age category, Finish time 1984 1 69 J Maitland Aberdeen ACC MOPEN 2:44:36 1984 2 53 B Brinfle Horwich RMI M50 2:50:36 1984 3 64 ARJ Curtis Livingston & D W50 2:52:34 1984 4 24 S Moore Horwich RMI M40 2:53:01 1984 5 65 AW Spenceley Carnethy HR WOPEN 2:56:55 1984 6 77 M Lindsay Carnethy HR MOPEN 2:58:42 data types data set type data availability attribute types ordering direction 19  Summary Data types: nature of the data (5) – items, attributes, links, positions, grids Data set types: how the data is arranged (4) – tables, networks, fields, geometry When the data is available (2) – static, dynamic Attributes: properties of the data (2) – categorical, ordered (ordinal, quantitative) Direction: ways of ordering (3) – sequential, diverging, cyclic 20 Data Types 21