Tableau walkthrough

  1. Open Tableau and select Connect>Text File.  Find your .csv file where you downloaded it (desktop or downloads, probably)

bar chart 23

2. Your data will take a second to upload, and it should display a preview of your data once it’s uploaded.  Select the tab for Sheet 1 at the bottom

bar chart 22

3. Begin by dragging two dimensions, two measures, or a dimension and a measure into columns and rows.  Use your comparisons from your proposal as a starting point, and think about what question you’re trying to answer.  Start simple with just one dimension/measure in each–for example, “what’s the relationship between race and trade?”

bar chart 21

4. Your data might not look like much at first!  Click through the examples in Show Me on the far right to see what kind of data each visualization needs–for example, a bar chart needs at least one measure, and if we’re looking at race and trade, we’ve given it two measures.  To create a measure (or create a dimension), click on the little downward arrow on the dimension or measure you want to change, hover over measure, and select Count from the menu.  Selecting measure tells Tableau to count the number of each item in that column, selecting dimension tells it to treat the information as a category, and selecting attribute tells Tableau to treat that information as true/false.

bar chart 20

5. You have a bar chart!  But what if you want more detail, ie, you want to see how many of each race/complexion practiced each trade?  Drag the dimension or measure you want more detail on to color in the Marks pane to further differentiate your data.

bar chart 19

6. Depending on your dataset, Tableau may as if you want to filter your data before adding it–in most cases, select Add all members, not Filter and then add.

bar chart 15

7. This will get you a bar chart with the bars sorted by race and colored by trade.  Yay!  But it’s very messy, and we want to have a very clear visual for our reader 🙁

bar chart 14

8. You have a couple of options for cleaning up your visual.  If you add another dimension or measure to your columns/rows, Tableau will group your data.  The order is important–whatever comes first is the order Tableau will order your data in, so think about what your important groupings are.  See below for complexion>trade vs trade>complexion.

bar chart 17

bar chart 18

8. However, you don’t have to clean up your data by grouping your columns–you can also simplify your color groups by minimizing the number of trades you display.  You may need to do this anyway if some of your data is misspelled or duplicated–for example, “baker” and “bqker” are the same trade, but Tableau doesn’t know that unless you tell it so, and “seaman,” “mariner,” “sailor” and “sailtor” are all the same trade.

You may need to do some thinking about why you’re grouping certain things together, or do some research on your kumquats.  For example, what’s a bellis maker?  Should a ribbon maker be grouped with weavers or with stay makers?  You can come back to this later, since this is probably where the majority of your research will be focused.  A good rule of thumb is under 20 groups–people have difficulty seeing differences between more than 20 groups.  But you don’t want to simplify to too few categories–it won’t make any sense if you have a category of “baker, sailor, apothocary, barber and farmer” and another category of “weaver, shoemaker, brewer, doctor” — what connects your categories together?

To group items, ctrl+click on the two or more items you want to group together, then right click and select Group.

bar chart 13

9. You may also want to Sort your data to make it more readable (this makes it easier for your reader to pick out the biggest and smallest groups).  Click the down arrow on the pane you would like to sort and select Sort.

bar chart 12

10. Select the method you would like to sort by–the default is data source order, but you may want to sort alphabetically or by the count of people in each trade or racial group.  Click Apply to preview the change, and OK to make it permanent.

bar chart 11

11. Depending on how you grouped and sorted your data, you might end up with categories sharing colors or two of the same color next to each other.  Click Color in the Marks pane to edit your colors.

bar chart 10

12. If you just hit OK immediately, nothing will happen–select Assign Palette to change the color designations, and then click ok to accept .  You can also manually choose colors, but this is a pain in the butt.  Think about the connotations your colors have for the story and argument you’re making–a light>dark color progression will imply that your data changes smoothly from one category to the next, which might not be the case if you’re looking at descriptive text!

bar chart 9

13. If the names of your groups get very long and unreadable, you may want to change them to make them more understandable.  You can do this two ways: right click the group and select edit alias, or right click the dimension in the left hand pane (make sure to select the (group) version!) and select aliases.  Both methods will bring up a window where you can change the text label that displays for that group.  Tableau will remember what your groups are, and you can see them by right clicking the grouped dimension and selecting Aliases.

bar chart 6

bar chart 8

14. You can also hide categories or group them together–click the category you want to hide and select exclude, or ctrl+click and select group.  Again, think carefully about what you’re hiding or grouping and why!

bar chart 5 bar chart 4

15. If you want to change any of the text formatting of your visualization, select Format in the toolbar.  Worksheet will change the defaults for all text, but you can select individual elements as well.

bar chart 3

16. Your final project requires at least two visualizations–if you want to give yourself a starting point without re-doing from step one, right click your sheet to duplicate it, and then play around with the Show Me options.  If you end up with something odd, refer back to our readings on different visualization types.

bar chart 2 bar chart 1

 

Maps in Tableau

  1. Follow steps 1-3 as above, but at step 3, you may need to first change the geographic role of one of your dimensions if Tableau is not already detecting it as geographic information (it will have a little globe next to it if it is).  Right click your dimension and select geographic role>country/region.  Depending on your research question, you may select other geographic levels, but unfortunately Tableau can’t handle more than one at a time.  Drag your geographic dimension into columns or rows and select the Filled Map option in Show Me if Tableau doesn’t generate a map automatically.

Screen Shot 2016-04-04 at 9.41.06 PM

2. To differentiate your countries, drag Birthplace into Color in the Marks pane.  Depending on your dataset, Tableau may as if you want to filter your data before adding it–in most cases, select Add all members, not Filter and then add.

Screen Shot 2016-04-04 at 9.45.38 PM

3.  If you’re working with census data, you will need to clean up your data to correspond to modern countries.  You may need to do some thinking about why you’re grouping certain things together, or do some research on your kumquats.  For example, where is Bayren, Prussia and Alsace?  Should Atlantic Ocean and Born at Sea be grouped together?  You can come back to this later, since this is probably where the majority of your research will be focused.

To group items, ctrl+click on the two or more items you want to group together, then right click and select Group.

Screen Shot 2016-04-04 at 9.49.46 PM

4. If you end up with quite a lot of places that need to be grouped together, group one or two in the colors list, and then in the dimensions list, right click the grouped dimension, which will bring up a larger pane to group your locations.

Screen Shot 2016-04-04 at 9.49.54 PM

5. To group your locations, crtl+click all the locations you want to group, and then rightclick>Group.  You should also rename your groups so that Tableau understands where each group is located.  Some outdated names, like Isle of Cuba or East India may need to be edited to their modern names.

Screen Shot 2016-04-04 at 10.01.52 PM

Screen Shot 2016-04-04 at 10.02.01 PM

6. To color your map by an attribute other than country name, drag the dimension you just grouped to detail rather than color–this keeps your countries selected but allows you to color them by another aspect.  Then drag the aspect you want to color your map with to color.  If you add a dimension to color, you may need to change it to a measure by right clicking and selecting measure>count (for example, if you want to color your map by the number of people born in that country).

Screen Shot 2016-04-04 at 10.11.38 PM

Screen Shot 2016-04-04 at 10.11.49 PM

7. Your map may not look like much at first–especially for census birthplace data, Tableau counts all the people born in the US or with no birthplace listed even if it doesn’t display them.  To color your map only by the countries displayed, right click your grouped dimension and select filter to exclude some locations from the count.

 

Screen Shot 2016-04-04 at 10.14.33 PM

8. By default, Tableau gives you a list of all the data included.  Deselect information you don’t want included in your count–for example, Null, which means no birthplace listed, or transcription errors like gender–by deselecting the check box for that category.  You can also reverse this by clicking Exclude at the bottom of the pane, which will allow you to select just the data you want excluded.

Screen Shot 2016-04-04 at 10.18.15 PM

9. You may want to edit your colors (see step 11 above).  A smooth progression of color can make it difficult to tell the difference between amounts–select stepped color and how many steps you want to make it easier for your reader to see the differences in amount.

Screen Shot 2016-04-04 at 10.21.47 PM