- You’ve already downloaded a data file.
- You remember where you put it. and
- You know what type of file it is (.csv or .xlsx.)
Step 1: Start Tableau Desktop & Connect to Your Data
If you have trouble finding your file in Tableau, double check that you selected the correct file type from the menu. For example if your data is a “.csv” , choose Text File. Otherwise, the file will be hidden from view in the file browser.
Overview of the Data Source Screen
Step 2: Check Your Data
The temptation to click the bright orange “Sheet 1” button and start making cool charts is almost irresistible. However, spending a little bit of time on the data upfront will save you time and prevent frustrations later.
- Get to know the data by looking at the column headings, -called field names in Tableau.
- Make sure you know what all of the fields are. If you don’t know some fields, find the data dictionary (sometimes called metadata or data definitions) and look them up.
- Rename the confusing fields. Do this after you look up data definitions, so you don’t lose any.
- Tableau guesses the data type, but sometimes gets it wrong. This is a good place to correct the data type if needed.
Fixing names and data types
Tableau will show the data in the first rows, if the data is small enough to easily preview. Look over the data in the first rows to see if it is what you expect it to be.
- If it doesn’t look nice and tidy, there could be something wrong with the source file delineation or otherwise corrupted.
- If numbers look weird (or are left padded with zeros), check and make sure Tableau knows they are numbers.
- If dates look totally wrong or are missing, check the data type. If the datatype is correct and dates still look crazy, then you may have an issue with date format in the source data.
- What is one record (or one row)? In relational databases one record is often one thing, like one patient encounter, or one day of COVID-19 cases. It is important to take note of this level of granularity.
At this point, Big League Tablusers* might also take the time to do other things on the Data Sources Screen. The one capability worth mentioning. -because you can NOT do it anywhere else in Tableau, is pivot. Pivot reshapes data that is wide (has lots of columns) and turns it into fewer columns but lots more rows.
*Pronounced tab-LOO-sers. (I’m just trying it out to see if it sticks.) It is a portmanteau of Tableau+Users. Portmanteau is French for “Squish two words together”.
Step 3: Save it!
If everything look OK, go ahead and save your project somewhere you will remember before clicking Sheet 1.