This lesson's interactive features are locked, please to keep using them

Data Types

Before you manipulate data, you need to know what kind of data you're looking at.

If you think a value is a number but it's actually a string, your calculations break. If you think a field always holds a string but sometimes it's None, your cleaning logic breaks. Here's a quick refresher on the Python types you'll most often encounter in data work:

int – whole numbers like 42
float – decimal numbers like 3.14
str – text like "hello"
bool – True or False
list – ordered collection of values
tuple – ordered collection that shouldn't change
set – unordered collection of unique values
dict – key-value pairs
NoneType – i.e. None, representing the absence of a value

Bad Types, Bad Calculations

Raw data almost never arrives in the exact type you want. For example:

movie = {"title": "Stranger Things", "rating": "8.7", "seasons": "4"}

Numbers as strings?! Unfortunately, this is common. If you're not careful, you might write:

boosted_rating: float = movie["rating"] + 0.5

Which will fail on a string. You need to convert the type first:

rating = float(movie["rating"])
boosted_rating: float = rating + 0.5

Assignment

SnackStack's kitchen sensors are sending readings with numbers stored as strings... classic appliance behavior.

Complete the clean_sensor_reading function. It accepts a dictionary of "raw" string-based sensor data and should return a new dictionary with the correct types.

The device_id should remain a string.
Convert temperature to a float.
Convert humidity to a float.
Convert battery_level to an int.
Convert online to a bool by checking whether its string value is "true" (the input will be "true" or "false").