Originally posted on Analytic Bridge
In Tableau, there are several data types that are supported. For example, you may have text values, date values, numerical values, and more. Each of the data types can take on different roles that dictate their behavior in the view.
All fields in a data source have a data type. The data type reflects the kind of information stored in that field, for example integers (410), dates (1/23/2005) and strings (“Wisconsin”).
Mixed Data Types for Excel and CSV Files
Most columns in an Excel or CSV (comma separated value) file contain values of the same data type (dates, numbers, text). When you connect to the file, Tableau creates a field in the appropriate area of the Data window for each column. Dates and text values are dimensions, and numbers are measures.
However, a column might have a mixture of data types such as numbers and text, or numbers and dates. When you connect to the file, the mixed-value column is mapped to a field with a single data type in Tableau. Therefore, a column that contains numbers and dates might be mapped as a measure or it might be mapped as a date dimension. The mapping is determined by the data types of the first 16 rows in the data source.
Example: if most of the first 16 rows are text values, then the entire column is mapped as text.
Empty cells also create mixed-value columns because their formatting is different from text, dates, or numbers.
Depending on the data type Tableau determines for each field, the field might contain Null values for the other (non matching) records as described in the table below.
You can read more about this at Tabeau Tutorial