Python中如何生成指定范围内符合函数输出的随机数据集？

Hey there! Let's break down how to generate that dataset you need. The core issue here is expanding the limited 0.0-1.0 range from random.random() to your full mymin-mymax interval, then applying your function to those scaled values. Here's how to do it:

Basic Approach with Python's Built-in random Module #

First, we'll map the 0-1 random values to your desired range using a simple formula:
x = mymin + random.random() * (mymax - mymin)

This formula takes the 0-1 float, scales it to fit the span of your range, then shifts it to start at mymin. Once we have our x in the right range, we just apply your timestwo(x) function to get the corresponding y value.

Here's a complete code example:

import random

# Your target function
def timestwo(x):
    return x * 2

# Function to generate the dataset
def generate_custom_dataset(mymin, mymax, num_points):
    dataset = []
    for _ in range(num_points):
        # Scale random 0-1 value to fit mymin-mymax range
        x = mymin + random.random() * (mymax - mymin)
        # Calculate y using your function
        y = timestwo(x)
        dataset.append((x, y))
    return dataset

# Example: Generate 100 points between 5 and 20
my_dataset = generate_custom_dataset(5, 20, 100)

Faster Alternative for Large Datasets with NumPy #

If you need to generate thousands or millions of points, using NumPy will be way more efficient than looping with the built-in random module. NumPy's np.random.uniform() lets you generate an array of values directly in your desired range:

import numpy as np

def timestwo(x):
    return x * 2

def generate_numpy_dataset(mymin, mymax, num_points):
    # Generate all x values at once in [mymin, mymax]
    x_values = np.random.uniform(low=mymin, high=mymax, size=num_points)
    # Calculate y values (works with arrays too!)
    y_values = timestwo(x_values)
    # Combine into a single dataset array
    return np.column_stack((x_values, y_values))

# Example: Generate 1000 points between 5 and 20
numpy_dataset = generate_numpy_dataset(5, 20, 1000)

Quick Note for Integer Values #

If your use case requires integer x values instead of floats, swap random.random() with random.randint(mymin, mymax) (for the built-in module) or np.random.randint(mymin, mymax+1, size=num_points) (for NumPy) to get whole numbers in your range.

内容的提问来源于stack exchange，提问作者sam