Embedded Files
Lesson 1 ❮ Lesson List ❮ Top Page
❯ 1.2 Data Types Conversion
⎺⎺⎺⎺⎺⎺⎺⎺⎺⎺⎺⎺⎺⎺⎺⎺⎺⎺⎺⎺⎺⎺
EXPECTED COMPLETION TIME
❲▹❳ Video 5m 37s
☷ Interactive readings 5m
Types of Data
Types of Data
pandas does not broadly classify data as either continuous or categorical.
The following table contains all pandas data types, with their string equivalents, and some notes on each type:
Common data
type name
Boolean
Integer
Float
Complex
Object
Datetime
Timedelta
Categorical
type name
Boolean
Integer
Float
Complex
Object
Datetime
Timedelta
Categorical
NumPy/pandas
object
np.bool
np.int
np.float
np.complex
np.object
np.datetime64
np.timedelta64
pd.Categorical
object
np.bool
np.int
np.float
np.complex
np.object
np.datetime64
np.timedelta64
pd.Categorical
Pandas string
name
bool
int
float
complex
O, object
datetime64
timedelta64
category
name
bool
int
float
complex
O, object
datetime64
timedelta64
category
Converting from numerical to categorical
Converting from numerical to categorical
To convert numerical to categorical, the astype method can be used. This method can cast a pandas object to a specified dtype.
> Note that the column facenumber_in_posters only consists of integer but pandas read it as float64. This is because the column has some missing values. Column with missing values cannot be directly converted to integer.
Converting from categorical to numerical
Converting from categorical to numerical
To convert categorical to numerical, the get_dummies method can be used. It will convert the categorical column into a 0-1 columns called dummy variables.
> It is possible to perform get_dummies for numerical column. It will generate a lot of columns and those might not be too useful.
Page updated
Google Sites
Report abuse