NUMPY Tutorial

🔹 NumPy Array Search

Searching within NumPy arrays is essential for filtering, locating elements, or conditional extraction. NumPy provides several useful functions to perform searches efficiently.

✅ Key Search Functions

• np.where(condition, [x, y]) — Returns indices or chooses values based on condition.
• np.argmax() — Returns the index of the maximum value.
• np.argmin() — Returns the index of the minimum value.
• np.nonzero() — Returns indices of non-zero elements.
• np.flatnonzero() — Returns indices of non-zero elements, flattened.
• np.searchsorted() — Finds indices to insert elements in sorted arrays.

🎯 Basic Syntax & Examples

import numpy as np

arr = np.array([10, 20, 30, 40, 50, 60, 70, 80])

# 1. np.where(): Find indices where condition is true
indices = np.where(arr > 35)
print("Indices where arr > 35:", indices)
# Output: (array([3, 4, 5, 6, 7]),)

# 2. np.where() with x, y parameters: Conditional replacement
new_arr = np.where(arr > 35, arr, 0)
print("Elements > 35 kept, others replaced by 0:", new_arr)
# Output: [ 0  0  0 40 50 60 70 80]

# 3. np.argmax() and np.argmin(): Index of max/min
max_index = np.argmax(arr)
min_index = np.argmin(arr)
print("Index of max value:", max_index)  # 7
print("Index of min value:", min_index)  # 0

# 4. np.nonzero(): Indices of non-zero elements
arr2 = np.array([0, 1, 0, 2, 3, 0])
nonzero_indices = np.nonzero(arr2)
print("Indices of non-zero elements:", nonzero_indices)
# Output: (array([1, 3, 4]),)

# 5. np.flatnonzero(): Flattened indices of non-zero elements
flat_nonzero = np.flatnonzero(arr2)
print("Flat non-zero indices:", flat_nonzero)
# Output: [1 3 4]

# 6. np.searchsorted(): Find insertion points to maintain order
sorted_arr = np.array([10, 20, 30, 40, 50])
insert_pos = np.searchsorted(sorted_arr, 35)
print("Position to insert 35:", insert_pos)  # 3

✅ Notes

• np.where(condition) returns a tuple of arrays (one for each dimension) with indices satisfying the condition.
• Use np.searchsorted() only on sorted arrays to find insertion points efficiently.
• np.nonzero() and np.flatnonzero() are useful to find all positions where elements are non-zero or True.

🔎 Summary

• Searching is done using conditional checks and index retrieval.
• Functions like where, argmax, argmin, nonzero, and searchsorted help with different search needs.