Arrays with NumPy - Learn Python with BNHR

NumPy¶

This notebook introduces NumPy (Numerical Python), the fundamental package for scientific computing in Python. It provides support for large, multi-dimensional arrays and matrices, along with a collection of mathematical functions to operate on these arrays.

Installation and Import¶

To use NumPy, we first need to import it. The community standard is to import it as np.

import numpy as np

print("NumPy version:", np.__version__)

Creating Arrays¶

Unlike Python lists, NumPy arrays are homogenous (all elements must be of the same type). This allows them to be much faster and more memory efficient.

Creating arrays from lists¶

We can create an array using np.array().

# A list of daily significantly completed surveys by a PSA enumerator
daily_surveys = [12, 15, 14, 10, 18]

# Convert to NumPy array
survey_array = np.array(daily_surveys)

print(survey_array)
print(type(survey_array))

Built-in array generators¶

NumPy offers functions to generate arrays automatically, which is useful for creating datasets for testing.

np.arange(start, stop, step): Like Python’s range but returns an array.
np.linspace(start, stop, num): Creates num evenly spaced values between start and stop.
np.zeros((rows, cols)): Creates an array filled with zeros.
np.ones((rows, cols)): Creates an array filled with ones.

# Create an array of Region IDs (0 to 9)
regions = np.arange(0, 10)
print("Regions (arange):", regions)

# Create 5 evenly spaced data points between 0 and 100
percentages = np.linspace(0, 100, 5)
print("Percentages (linspace):", percentages)

# Create a 3x3 matrix of zeros (placeholder for missing data)
missing_data = np.zeros((3, 3))
print("\nZero Matrix:\n", missing_data)

Array Attributes¶

Inspect the structure of your data using these attributes:

ndim: Number of dimensions (axes).
shape: The dimensions of the array (rows, columns).
size: Total number of elements.
dtype: The data type of the elements (e.g., int64, float64).

# 2D Array representing population counts in 2 Districts over 3 years
# Row = District, Column = Year
pop_data = np.array([
    [1500, 1600, 1700], 
    [2000, 2100, 2200]
])

print("Dimensions (ndim):", pop_data.ndim)
print("Shape (rows, cols):", pop_data.shape)
print("Total Elements (size):", pop_data.size)
print("Data Type (dtype):", pop_data.dtype)

Practice Exercises¶

The following exercises are designed to test your understanding of NumPy basics.

Creating a Household Income Array¶

Task: Create a NumPy array from a list of household incomes provided below. Print the array and its data type.

# List of household incomes in PHP
incomes_list = [25000, 32000, 18000, 45000, 29000]

# Write your solution here

Income Array: [25000 32000 18000 45000 29000]
Data Type: int64

Generating ID Codes¶

Task: Use np.arange to generate an array of ID codes from 100 to 110 (inclusive of 100, exclusive of 110).

# Write your solution here

ID Codes: [100 101 102 103 104 105 106 107 108 109]

Placeholder Data¶

Task: Create a 1D array of 5 zeros using np.zeros to serve as a placeholder for incoming census data.

# Write your solution here

Placeholders: [0. 0. 0. 0. 0.]

Reshaping Data¶

Task: Create a 1D array containing 12 data points representing monthly inflation rates. Reshape this array into a 2D array with 4 rows (Quarters) and 3 columns (Months).

# Write your solution here

Quarterly View:
 [[ 0  1  2]
 [ 3  4  5]
 [ 6  7  8]
 [ 9 10 11]]

Broadcasting Operations¶

Task: You have an array of population counts. You to project a 10% growth for next year. Multiply the entire array by 1.10.

current_pop = np.array([1000, 2500, 3000, 4200])

# Write your solution here

Projected Population: [1100. 2750. 3300. 4620.]

Statistical Calculations¶

Task: Given a list of age data, calculate the mean (average) age and the standard deviation.

age_list = [23, 45, 18, 60, 34, 29, 51]

# Write your solution here

Average Age: 37.14
Standard Deviation: 14.24

Filtering and Sorting¶

Task 1: Filter an array of city populations to find only those cities with a population greater than 1000000.

# Write your solution here

pop_dict = [
    {"name": "Manila", "pop2025": 1600000},
    {"name": "Quezon City", "pop2025": 2900000},
    {"name": "Cebu City", "pop2025": 922300},
    {"name": "Davao City", "pop2025": 1832753},
    {"name": "Zamboanga City", "pop2025": 985595},
    {"name": "Antipolo", "pop2025": 785715},
    {"name": "Pasig City", "pop2025": 755200},
    {"name": "Taguig City", "pop2025": 804442},
    {"name": "Cagayan de Oro City", "pop2025": 675555},
    {"name": "Iloilo City", "pop2025": 457358}
]

# Solution
pop_array = np.array(pop_dict)

# Filter the array to find only those cities with a population greater than 1000000
filtered_pop = pop_array[np.array([city['pop2025'] > 1000000 for city in pop_dict])]

print(filtered_pop)

[{'name': 'Manila', 'pop2025': 1600000}
 {'name': 'Quezon City', 'pop2025': 2900000}
 {'name': 'Davao City', 'pop2025': 1832753}]

Task 2: What are the top 5 cities based on population? Return only the names of the cities.

# Write your solution here

#Solution
pop_array = np.array(pop_dict)

# Convert the 'pop2025' value into a numpy array
pops = np.array([city['pop2025'] for city in pop_dict])
cities = np.array([city['name'] for city in pop_dict])

top_5_indices = np.argsort(pops)[::-1][:5]

print("The top 5 cities in terms of population are")
for city in top_5_indices:
    print(cities[city])

The top 5 cities in terms of population are
Quezon City
Davao City
Manila
Zamboanga City
Cebu City