Summary - Performance
- Python course
- Expert
- Lesson 42/47
- 5 min
Combine timeit, cProfile and __slots__ to write an optimized handler.
Recap: Performance
In this group you learned Python performance tools. You measured with timeit and profiled with cProfile.
You measured memory with sys.getsizeof. You saved memory with __slots__. You applied vectorization with numpy. Now you will combine these into an optimized handler: a memory-friendly class with __slots__ and fast computation with numpy together.
Practice Write an optimized handler
Import the module with import numpy as np. Write a Point class with __slots__ = ['data']. In __init__(self, data) set self.data = np.array(data). Add a squared method returning self.data ** 2. Create a Point([1, 2, 3]) instance and print .squared().
Starter code
# write the Point class: slots + numpy
# create an instance and printSample solution
import numpy as np
class Point:
__slots__ = ["data"]
def __init__(self, data):
self.data = np.array(data)
def squared(self):
return self.data ** 2
p = Point([1, 2, 3])
print(p.squared())Expected output
[1 4 9]
What you learned
In this group you learned Python performance optimization. You measured with timeit and profiled with cProfile. You measured memory with sys.getsizeof.
You saved memory with __slots__ and applied vectorization with numpy. In the summary you combined these into an optimized handler. This knowledge is critical for large-data and high-performance applications.
Questions and answers
Review the lesson questions together with their answers. To test yourself, solve the lesson interactively.
1 Which tool for which job?
- timeit to measure memory
- getsizeof for vectorization
- cProfile for time profiling Correct answer
- numpy for profiling
Explanation: cProfile profiles function times. The other matches are wrong.
2 Match all performance concepts
| Term | Meaning |
|---|---|
timeit |
Measures time |
cProfile |
Profiles |
getsizeof |
Measures memory |
numpy |
Vectorizes |
3 Final blank
import numpy as ____
a = np.array([1, 2, 3])import numpy as np
a = np.array([1, 2, 3])Explanation: import numpy as np is the standard alias.