Data type constraints

Cleaning Data in Python

Adel Nehme

Content Developer @ DataCamp

Course outline

dirty_data

Course outline

side effects

Course outline

clean_data

Course outline

clean_data

Chapter 1 - Common data problems

Why do we need to clean data?

ds_workflow

Why do we need to clean data?

ds_workflow

Why do we need to clean data?

Garbage in Garbage out

Data type constraints

Datatype	Example
Text data	First name, last name, address ...
Integers	# Subscribers, # products sold ...
Decimals	Temperature, $ exchange rates ...
Binary	Is married, new customer, yes/no, ...
Dates	Order dates, ship dates ...
Categories	Marriage status, gender ...

Python data type
`str`
`int`
`float`
`bool`
`datetime`
`category`

Strings to integers

# Import CSV file and output header
sales = pd.read_csv('sales.csv')
sales.head(2)

   SalesOrderID    Revenue    Quantity
0         43659     23153$          12
1         43660      1457$           2

# Get data types of columns
sales.dtypes

SalesOrderID    int64
Revenue         object
Quantity        int64
dtype: object

String to integers

# Get DataFrame information
sales.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 31465 entries, 0 to 31464
Data columns (total 3 columns):
SalesOrderID     31465 non-null int64
Revenue          31465 non-null object
Quantity         31465 non-null int64
dtypes: int64(2), object(1)
memory usage: 737.5+ KB

String to integers

# Print sum of all Revenue column
sales['Revenue'].sum()

'23153$1457$36865$32474$472$27510$16158$5694$6876$40487$807$6893$9153$6895$4216..

# Remove $ from Revenue column
sales['Revenue'] = sales['Revenue'].str.strip('$')
sales['Revenue'] = sales['Revenue'].astype('int')

# Verify that Revenue is now an integer
assert sales['Revenue'].dtype == 'int'

The assert statement

# This will pass
assert 1+1 == 2

# This will not pass
assert 1+1 == 3

AssertionError                            Traceback (most recent call last)
         assert 1+1 == 3
AssertionError:

Numeric or categorical?

...   marriage_status    ...
...                 3    ...
...                 1    ...
...                 2    ...

0 = Never married 1 = Married 2 = Separated 3 = Divorced

df['marriage_status'].describe()

       marriage_status
...
mean              1.4
std               0.20
min               0.00
50%               1.8 ...

Numeric or categorical?

# Convert to categorical
df["marriage_status"] = df["marriage_status"].astype('category')

df.describe()

        marriage_status
count                 241
unique                4
top                   1
freq                  120

Let's practice!

Cleaning Data in Python