Tutorials
From INLS507_Resources
Contents[hide] |
[edit] Publishing in Tableau
Create worksheet as you normally would. Then in top menu select
Server>PublishWorkBook -or- TableauPublic>SaveToWeb
and save your result to Tableau Public (both options do this; you'll need to have setup a Tableau public account).
It will also preview this on your screen in popup window. (or you can login to Tableau Public under your account and access). On either of these windows showing your tableau public worksheet,
CLICK ON the SHARE label in bottom left (next to Facebook/Twitter icons).
Clicking on this will give you URL you can use to provide others access to your "public" worksheet. See for instance one I put up:
http://public.tableausoftware.com/views/pub1/Sheet2?:embed=y
[edit] Tableau On-Demand Videos
Amazing Things To Do With Tableau
Ratings: 1 = not useful, 3=somewhat useful, 5= extremely useful
Recommend Starting With These:
- Explore the Interface (4:46), Rating:5 -- overview of layout, including introduction of shelves and menu options
- Connect to Excel (3:22), Rating:4 -- basics of uploading excel data, including multiple sheets
- Show me a Visualization (1:42), Rating:4 -- quick demonstration of show me tool
- Create a Visualization (4:02), Rating:5 -- shows drag and drop of measures and dimensions into shelves
- Put Data on Maps (4:53), Rating:4 -- introduction to mapping, including how to generate and map options
Other Very Useful Clips:
- Drill into Data (2:37), Rating:5 -- how to view underlying data for select marks
- Filter Data (4:20), Rating:5 -- filter shelf, quick filters, and filtering criteria
- Analysis of Time Series (3:01), Rating:4 -- viewing data with dates and using date parts
- Calculating Changes (4:38), Rating:5 -- using quick table calculations (e.g., percent difference)
- Group Data into Categories (4:35), Rating:5 -- ways to create new groups from data, including naming and ungrouping
- Create a Dashboard (4:00), Rating:5 -- putting multiple views together, highlighting and filtering across views
- Customize Views (7:15), Rating:5 -- customizing titles and captions, changing the view/size, choosing color palettes
- Guided Analytics (6:59), Rating:5 -- highlighting, filtering, adding actions
And the Rest:
- Connect to SQL (2:49), Rating: 2 -- uploading multiple SQL tables and exporting data
- Working Offline (1:51), Rating:1 -- using extract to work offline
- Add Quadrants and Thresholds (3:29), Rating:3 -- using reference lines to create quadrants and thresholds
- Annotate Findings (3:49), Rating:3 -- annotating data marks, points, and areas
- Link to the Web (2:03), Rating:2 -- embed websites within Tableau
- Share on the Web (3:51), Rating:1 -- publishing a workbook via a tableau server
- Share with Free Reader (3:10), Rating: 3 -- viewing packaged workbooks with the free Tableau reader
- Cut and Paste Data (2:15), Rating:3 -- cut and paste data from word documents, websites, etc. into Tableau
- Numbers (3:21), Rating:3 -- ways to get at the hard numbers: annotation, data labels, underlying data
[edit] Tableau and Salary Dataset: Getting Started Instructions
Assuming Tableau is installed, open Tableau and start a new workbook.
Data> Connect to Data and select Microsoft Excel (or appropriate) and browse to pick your datafile. Tableau will then prompt to create an extract from excel to improve performance. You can save the connection to the data to reuse later.
Note: before connecting you'll need to remove any extra information (rows) from your dataset. You should only have the data plus headers (top row). (If you forget this, once columns/rows are displayed in the Tableau worksheet, right click on extraneous data and select EXCLUDE. Also, if your header names aren't recognized, you can rename by right clicking the dimension or measure.)
Now, you should see things appear in the Data column on the left under Dimensions and Measures. Dimensions are our "nominal, ordinal" sets. Their measures are our "continuous" data type. Depending on how you are trying to visualize the data, you may need to change a dimension into a measure or vice versa (just drag and drop into the appropriate column). You can also reclassify the data type as needed (right click and scroll down to Change Data Type).
You can then drag and drop measures and dimensions as indicated to create columns and rows. Or you can select several dimensions/measures and click the Show Me! tab to choose a data view. Measures/dimensions can also be dropped onto different shelves (e.g., pages, filters, marks) to manipulate the visualization.
To create another sheet within the same workbook, right click on the sheet tab. You can also make a small multiples visualization by selecting "New Dashboard" and then dragging several visualizations (their sheets) into one display area (dashboard).
[edit] Doing Maps with our Tools
How to do Map based visualizations with the big three (Excel/MapPoint, ManyEyes, Tableau).
We describe how to do A7 (map based visualizations with the three commonly used tools that can intrepret data values for states etc, and show on visual map of US/World.
[edit] Tableau and Olympics Dataset
See above (Tableau and salary) for importing steps.
This time you should see a small sphere "globe" to left of name of attribute that has map data (e.g., country name, county name, US state, etc.).
- to create a map you can either just double click on the geographical dimension
- OR you can drag latitude and longitude from measures. Note that latitude and longitude are generated automatically from geographical names and are labeled as such ("generated").
- circles will appear for each geographical location given in the dimension data
To visualize medal count:
- to show medal count by size drag measure onto the Size shelf. This will change the size of the geographical circle accordingly. You can adjust the overall mark size by adjusting the slider under Size and you can adjust the mark scale under the Measure Name by rightclicking to "Edit Sizes."
To see multiple variables as a pie chart on a map:
- generate map from geographical dimension
- drag measure names (as a dimension) onto Color shelf
- drag measure values (as a measure)onto Size shelf
- Measure Values shelf will appear; remove any measure items you don't want to display
- adjust size and color transparency of pie marks by using the sliders under the shelves
To visualize multiple variables as different colored shapes on a map:
- generate map from geographical dimension
- in the Marks shelf drop down menu, select Shape
- drag measure names (as a dimension) onto Shape shelf
- also drag measure names (as a dimension) onto Color shelf
- drag measure values (as a measure) onto Size shelf
- to change Shape icons, left click on Measure Names icon shelf and Edit Shapes. Select shape palette from drop down menu, assign to data items, then apply to map. See below for info on uploading your own icons.
- to change Color scheme, left click on Measure Names color shelf and Edit Color. Select color palette from drop down menu, assign to data items, then apply to map
- to change Size scale, left click on Measure Values size shelf and Edit Sizes. Set size variation, range, and legend values.
To upload your own icons, see Tableau Online Help and search
- Custom Shapes
- Tips for Creating Custom Shapes
[edit] ManyEyes and Olympics Dataset
- Go to [http://manyeyes.alphaworks.ibm.com/manyeyes/ many eyes webpage
- you will need to create an account
- under participate on left menu select Upload Data Set
- separately you open up the file in Excel or similar and copy to the clipboard (^C) all the data you want (including header lables)
- then in manyeyes you paste into their window as free text
- then check that maneyes understood data (it shows you a snippet)
- you can manually edit the dataset within many eyes if needed
- fill in what you want for "tell us about your data" (this is for other people, note: all data/visualizations publicly viewable)
- click on visualize icon at bottom of your dataset
- you'll see a visual menu list of visualization types
- Select WORLD MAP. Some country maps and a US county map are also available.
- It will ask you what column/attribute contains the geographical data. For this dataset, put in "country".
- ignore all the unrecognized countries (don't worry this).
- manyeyes will give you a few fixed choices. So, look on the right side of menu list for "colors or bubbles". You can also choose 1/2/All for 1 or two or small multiples visualizations.
- Then you have some control at the bottom to change parameters, i.e. like whether you're showing gold/silver/bronze/or total medals.
Issues: Manyeyes map facility seems to only support one data parameter per geographic location (Country, state, county) at a time. YOu can select different ones, but you can't visualize two or three or more.
[edit] MapPoint and Olympics Dataset
- must download the trial version of MapPoint
- bring up MapPoint (which shows up as it's own program)
- Note that MapPoint seems to want to force you to just look at a portion of the world through their windowshield icon on top of the world map in the upper left corner.
- Load in your map. DATA>IMPORT DATA WIZARD> selected the file to load.
- Then it takes you to Data Mapping Wizard.
- Choose a visualization type
- then you'll choose data to map to the visualization technique (this will depend on your choice of visualization type).
- you then get formatting control to select size of things, etc.
(IF you already have a data file made from before you can just do FILE> OPEN> browse to load your map file).
[edit] Google Motion Charts
Follow these directions for uploading data into a Google doc and inserting a motion chart: http://www.gapminder.org/upload-data/motion-chart/
NOTE: You will need to transpose your data so it is in columns across (country, year, variable 1, variable 2). See my example dataset
Here are detailed steps to formatting your data:
- If the dataset is large, you will probably want to select just a subset to make the transposing process more manageable. Before transposing the data, I would recommend deleting rows for countries of lesser interest. You may also want to limit the time range by deleting some of the year columns.
- Once parsed, copy your entire dataset and open a new excel sheet. Rightclick and select Paste Special, check Transpose. This will flip the rows and columns. You should now have years in Column A and country-specific data in the remaining columns.
- You will need to create a new column A for country data. Paste in each country name for the number of corresponding data rows. That is, if you have 10 years of data, in column A the name for country #1 should be pasted into rows 2-11, the name for country #2 in rows 12-21, etc.
- Column B should now be years. Copy the years going down the rows so that each country in column A matches up to the same set of years. For example, years 2000-2010 for country #1 is in column B: rows 2-11, years 2000-2010 for country #2 is in column B: rows 12-21, etc.
- Column C is your first variable. Cut and paste the data (currently in columns across) for each country/years into the appropriate rows going down. For example, data for years 2000-2010 for country #1 is in column C: rows 2-11, data for years 2000-2010 for country #2 is in column c: rows 12-21, etc.
- Column D is your second variable pulled from another dataset. For example if you are using GDP data, you will need to download that excel sheet, transpose the data, and then paste in to Column D to match up with the correct country/years. For example, GDP data for years 2000-2010 for country #1 is in column D: rows 2-11, GDP data for years 2000-2010 for country #2 is in column D: rows 12-21, etc.
- Once your data is in the right format, cut and paste it into a Google Doc spreadsheet.
- Insert Gadget: Motion Chart, change the settings if desired (e.g., color), and play!
If you have any questions about any of the tutorial instructions, feel free to email me at juliatermaat@gmail.com.
