LabTwo: Interactive Data Exploration in MR

The LabTwo task involves the creation of a mixed reality application that allows a user to explore a data-set using an interactive bar-chart system. The concept of interactive data exploration is based on Taher et al.'s work "Exploring Interactions with Physically Dynamic Bar Charts" which was carried out at SCC. Link to the paper and a demonstration video are provided at the end of this document.

DataSets

Two datasets are provided:

1. UK Temperature Dataset: Monthly average maximum temperature in UK for last 60 years. Filename: UK_Max_Temp_60Yrs.csv Data structure: Columns: Year, Months: Jan-Dec, Seasons: Win-Aut, Annual Display requirements: X-axis > Year, at start most recent 12 years; Z-axis > Months, at start Jan-Dec, Y-axis > Temperature C Data point annotation: (what should be displayed for the annotation about the data point) Period: {mon/season/annual}, Year: {year}, Temperature: {value} C.

2. 20 Country CO2 Emission Dataset: Yearly emissions of CO2 by country for the last 72 years. Filename: CO2_Emissions_20_Country_72Yrs.csv Data structure: Columns: Country, Year, ISO_code (for country), Population_in_Million, GDP_Billion_USD, Pct_Share_Global_CO2, CO2_Million_tonnes, Cumulative_CO2_Million_tonnes Display requirements: X-axis > Year, at start most recent 12 years; Z-axis > Country, at start countries with highest cumulative CO2 values; Y-axis > CO2_Million_tonnes; Data point annotation: (what should be displayed for the annotation about the data point) Emissions for {Country}, in {Year}: {CO2_Million_tonnes} mi, Global Share: {Pct_Share_Global_CO2}% Population: {Population_in_Million}, GDP: ${GDP_Billion_USD} bi. Cumulative: {Cumulative_CO2_Million_tonnes} mi

Scenes

1. Data Selector Scene: Shown when application loads. User can see two data-sets as UI elements and the user can select one of them. When a dataset is selected, subsequent scene loads the data and prepares it for display. This can be rolled up into an UI element, instead of a separate scene, but it is not visible at the same time as the Interactive Bar Chart.
2. Interactive Bar Chart Scene: This scene is the main scene. The main elements of this scene are: a. hidden context menu, b. interactive bar chart. The design of each are described next

Hidden Context Menu

Contents: This menu consist of a button that takes the user back to the Data Selector Scene and another button which displays a general help UI panel. The general help UI panel displays a simplified description of how to interact with the bar graph. Consider how you are going to make this user friendly to read.

Interactive Bar Chart

The interactive bar chart consists of:

1. Floor which functions as a base for all the interactive elements
2. A 12 x 12 grid of interactive bars, the bar colour-scheme is consistent across the X-axis (e.g. all years for same country are coloured same)
3. Interactive label elements: 12 on X-axis and 12 on Z-axis, label element matches what is being displayed. (e.g. z-axis has months for the temperature dataset)
4. Two scroll-pads: one for each axis, allowing users to see different dataframes (each dataframe is 12 steps wide)
5. Y-axis scale bar showing maximum value. Y-axis "scale" depends on data being displayed. However, the maximum world height of these bars is limited to
6. Annotation label initially shows "Touch to select data-point", when the interaction results in data-point being selected, associated annotation as per specification provided for each dataset.

What should the user be able to do?

Data exploration capabilities

The system should allow the user to explore the data by navigating through it, check individual values, compare two series (e.g. two countries across years).

Action users should be able to perform

1. Show data-tuple of a single datapoint: Interact with a single datapoint directly on the grid and the annotation label will be updated to its value. See individual dataset specification under "Data point annotation" to see what should be shown.
2. Show data-tuples across an axis: Interact with an interactive label on one axis with one hand and simultaneously interacting with interactive labels on the other axis will result in the intersecting datapoint being highlighted and the annotation label will be updated with data associated with the datapoint.
3. Hide datapoint: Interact with a datapoint to hide it from view.
4. Show datapoint: Interact with the base position of a hidden datapoint to make it visible again.
5. Hide data-series: Interact with a data-series label to hide the entire data-series
6. Show data-series: Interact with a data-series label of a hidden data-series to make it visible again
7. Swap position of two data-series: Interact with two data-series labels to swap their position in the grid
8. Compare two data-series: Interact with two data-series labels to show those two data-series only and hide the remaining data-series.
9. Step through data-frames: Step through data-frames by interacting with the two scroll-pads. Hidden data-series should stay hidden

Information Sources:

Paper Information

URL: https://doi.org/10.1145/2702123.2702604 Youtube Link: https://www.youtube.com/watch?v=UC6l6dy04zI

DataSet Sources:

1. UK Met: https://www.metoffice.gov.uk/research/climate/maps-and-data/uk-and-regional-series
2. CO2 Emissions: https://ourworldindata.org/co2-emissions