By the end of this lesson, you can:

• Build a new dictionary by choosing a key and value for each input item.
• Filter entries and transform keys or values within a dictionary comprehension.
• Build a set comprehension to produce unique transformed values.
• Recognize that dictionary keys are unique and the last value wins when generated keys collide.

In the previous lesson, a list comprehension created a list from an iterable. Dictionary comprehensions use curly braces with a key: value pair:

new_dict = {key_expression: value_expression for item in source}

Use .items() to process dictionary keys and values together; add a trailing if to exclude entries. A dictionary keeps one value per key. If multiple input entries produce the same key, the later result replaces the earlier value. The source collection is not changed.

A set comprehension also uses curly braces, but has no colon:

new_set = {expression for item in source}

Sets retain unique values, but their iteration order is not guaranteed. When output needs a predictable order, sort the resulting set before printing or returning it as a list. As with list comprehensions, choose a regular loop if the transformation has several steps or side effects.

1. Turn existing values into a lookup

Choose a key and a derived value for each item:

packages = ["api", "worker", "mailer"]
name_lengths = {package: len(package) for package in packages}
print(name_lengths)

Output:

{'api': 3, 'worker': 6, 'mailer': 6}

Every package becomes a key; its length becomes the value.

2. Filter and adjust a dictionary

Use .items() to access both parts of each entry:

balances = {"ops": 55, "research": 20, "support": 80}
raised = {team: amount + 10 for team, amount in balances.items() if amount >= 50}
print(raised)

Output:

{'ops': 65, 'support': 90}

The research entry is excluded, and the original dictionary still contains its original values.

3. Normalize and deduplicate values

A set comprehension keeps each distinct cleaned label once:

raw_labels = [" Backend ", "backend", "DATA", "data "]
labels = {value.strip().lower() for value in raw_labels}
print(sorted(labels))

Output:

['backend', 'data']

We use sorted() for predictable display—not because the set itself has a sorted order.

4. Understand duplicate generated keys

When different values generate the same key, the later assignment wins:

entries = ["DEV", "dev", "Test"]
last_seen = {name.lower(): name for name in entries}
print(last_seen)

Output:

{'dev': 'dev', 'test': 'Test'}

Both first entries generate dev; the later original value overwrites the earlier one.

• A set comprehension does not keep duplicates or guarantee iteration order.
• if filters come after the for part, not before it.
• Empty input or no matching items produces {} for a dictionary or set() for a set.
• Normalizing different dictionary keys to the same text can overwrite earlier values.