NumPy Array From Existing Data – Copy, View, and Transform Arrays in Python

🐍 NumPy – Array From Existing Data

In NumPy, you don’t always create arrays from scratch.

Often, you work with existing data and build new arrays from it.

NumPy provides powerful tools to:

• Copy arrays
• Create views
• Slice data
• Reuse memory efficiently
• Transform existing datasets

These techniques are essential in data science and machine learning using Python.

---

What Does “Array From Existing Data” Mean?

It means creating a new array based on already existing NumPy arrays or Python data structures.

Instead of creating new values, you reuse or transform existing data.

---

🟢 1. Creating Arrays from Python Lists

import numpy as np

list_data = [1, 2, 3, 4]

arr = np.array(list_data)

print(arr)

---

🟢 2. Copy vs View (Important Concept)

Copy

A copy creates a completely new array.

arr = np.array([1, 2, 3])

copy_arr = arr.copy()

copy_arr[0] = 99

print(arr)
print(copy_arr)

Output:

[1 2 3]
[99 2 3]

✔ Original array is unchanged.

---

View

A view shares the same memory.

arr = np.array([1, 2, 3])

view_arr = arr.view()

view_arr[0] = 99

print(arr)

Output:

[99 2 3]

✔ Both arrays are linked.

---

🟡 3. Slicing Existing Arrays

Slicing creates a view of the original array.

arr = np.array([10, 20, 30, 40, 50])

new_arr = arr[1:4]

print(new_arr)

---

🟡 4. Reshaping Existing Data

arr = np.arange(6)

new_arr = arr.reshape(2, 3)

print(new_arr)

---

🟡 5. Flattening Existing Arrays

arr = np.array([[1, 2], [3, 4]])

flat = arr.flatten()

print(flat)

---

🔵 6. Joining Existing Arrays

Concatenate

a = np.array([1, 2])
b = np.array([3, 4])

print(np.concatenate((a, b)))

---

Stack Arrays

print(np.vstack((a, b)))
print(np.hstack((a, b)))

---

🔵 7. Splitting Existing Arrays

arr = np.array([1, 2, 3, 4, 5, 6])

parts = np.array_split(arr, 3)

print(parts)

---

🟣 8. Extracting Data from 2D Arrays

arr = np.array([
    [1, 2, 3],
    [4, 5, 6]
])

print(arr[0])
print(arr[:, 1])

---

🟣 9. Boolean Filtering from Existing Data

arr = np.array([10, 20, 30, 40, 50])

filtered = arr[arr > 25]

print(filtered)

Output:

[30 40 50]

---

🧠 Copy vs View Summary

Feature	Copy	View
Memory	Separate	Shared
Performance	Slower	Faster
Changes affect original	No	Yes
Use case	Safe editing	Efficient processing

---

⚡ Why This Concept is Important

1. Memory Efficiency

Views reduce memory usage.

2. Faster Computation

Avoid unnecessary duplication.

3. Data Processing

Used in filtering and preprocessing.

4. Machine Learning

Efficient dataset handling.

---

📊 Real-World Example

Dataset Processing

import numpy as np

data = np.array([100, 200, 300, 400, 500])

train_data = data[:3]

print(train_data)

Used in:

• Training datasets
• Validation sets
• Data slicing

---

🚀 Best Practices

• Use .copy() when you need safe data modification
• Use .view() for performance optimization
• Use slicing for quick data extraction
• Avoid unnecessary copies in large datasets
• Always understand memory behavior

---

🧾 Summary

NumPy allows you to create arrays from existing data efficiently using:

• Copy
• View
• Slicing
• Reshaping
• Joining
• Splitting
• Filtering

These techniques are essential for handling large datasets in Python.

---

🏁 Conclusion

Working with arrays from existing data is a powerful feature of NumPy.

It improves performance, reduces memory usage, and simplifies data manipulation.

Mastering these concepts is essential for data science, machine learning, and real-world Python applications.