Tabular data are the most common form of structured
data that we use for analysis in data science.
But what are tabular data and how do we organize
our data in this way?
Tabular data are data organized into a table.
The table provides the data with structure.
A table, is a two-dimensional grid of data.
However, unlike a matrix, which we saw earlier,
all of the elements in a table do not need
to be all of the same data type
Rather, all data in each column must be the
same data type, which we refer to as homogenous
However, all data in a row can have different
data types, from column to column, which we
refer to as heterogenous data.
For example, imagine we have a table of patients
at a hospital.
We would have a set of rows (one for each
patient) and a set of columns, (one for each
attribute of the patient).
Each element of data in a column must be the
same data type.
For example, – all of the names must be character
– all of the genders must be enumerations
of male, female, or other genders.
– all ages must be integers…
– and so one.
However, each row contains elements of various
– the name “Bill” is a character string,
– the gender “Male” is an enumeration
– and the age “21” is an integer
As we can see, each column contains only a
single data type; however, each row can contain
multiple data types.
In data science, tabular data can be broken
down into three main components:
– Observations – which we locate on the
rows of a table
– Variables – which we locate on the columns
of a table
– and Relationships – which connect data
in one table to data in another table
We’ll discuss each of these components, in
more detail, next.