如何在R或Python中生成指定均值与范围的100个随机值？

Got it! Let's walk through how to generate 100 random numbers with a specific mean and range in both R and Python. I'll share two practical approaches for each language—one that's quick and intuitive, and another that gives you a more natural distribution.

Approach 1: Adjust Uniform Random Numbers to Target Mean #

This is a simple, no-external-packages method. We start with uniform numbers across your desired range, then shift them to hit your target mean (while keeping values within the range).

# Define your parameters upfront
min_val <- 0    # Lower bound of your range
max_val <- 10   # Upper bound of your range
target_mean <- 5.5  # Desired mean
n <- 100        # Number of random numbers

# Step 1: Generate initial uniform random numbers
set.seed(123) # Use this for reproducible results
rand_nums <- runif(n, min = min_val, max = max_val)

# Step 2: Calculate how much we need to shift the numbers
current_mean <- mean(rand_nums)
adjustment <- target_mean - current_mean

# Step 3: Adjust the numbers and clamp to stay within your range
adjusted_nums <- rand_nums + adjustment
adjusted_nums <- pmax(min_val, pmin(max_val, adjusted_nums))

# Check the results
cat("Adjusted mean:", round(mean(adjusted_nums), 2), "\n")
cat("Min value:", round(min(adjusted_nums), 2), "\n")
cat("Max value:", round(max(adjusted_nums), 2), "\n")

Note: If your adjustment is large, some numbers might get clamped to the min/max boundaries, which can slightly skew the distribution. For a smoother distribution, try the next approach.

Approach 2: Truncated Normal Distribution #

For a more natural-looking distribution (closer to bell-shaped), use the truncnorm package to generate numbers that stay within your range and hit your target mean.

# Install the package first if you haven't
# install.packages("truncnorm")
library(truncnorm)

# Define parameters
min_val <- 0
max_val <- 10
target_mean <- 5.5
n <- 100

# Generate truncated normal random numbers
# Adjust the `sd` parameter to control how spread out the numbers are
set.seed(123)
rand_nums <- rtruncnorm(n, a = min_val, b = max_val, mean = target_mean, sd = 2)

# Verify results
cat("Mean:", round(mean(rand_nums), 2), "\n")
cat("Min:", round(min(rand_nums), 2), "\n")
cat("Max:", round(max(rand_nums), 2), "\n")

Tip: If the generated mean isn't exactly your target, tweak the sd value slightly—smaller sd keeps numbers closer to the mean, larger sd spreads them out more.

Approach 1: Adjust Uniform Random Numbers to Target Mean #

Just like in R, this method uses NumPy to generate uniform numbers, then shifts them to hit your target mean while respecting the range.

import numpy as np

# Define your parameters
min_val = 0    # Lower range bound
max_val = 10   # Upper range bound
target_mean = 5.5  # Desired mean
n = 100        # Number of random numbers

# Step 1: Generate initial uniform random numbers
np.random.seed(123) # Reproducible results
rand_nums = np.random.uniform(min_val, max_val, n)

# Step 2: Calculate adjustment factor
current_mean = rand_nums.mean()
adjustment = target_mean - current_mean

# Step 3: Adjust numbers and clamp to range
adjusted_nums = rand_nums + adjustment
adjusted_nums = np.clip(adjusted_nums, min_val, max_val)

# Check results
print(f"Adjusted mean: {adjusted_nums.mean():.2f}")
print(f"Min value: {adjusted_nums.min():.2f}")
print(f"Max value: {adjusted_nums.max():.2f}")

Approach 2: Truncated Normal Distribution with SciPy #

Use SciPy's truncnorm to generate bell-shaped random numbers within your range. We'll calculate the necessary parameters to align with your target mean.

from scipy.stats import truncnorm
import numpy as np

# Define parameters
min_val = 0
max_val = 10
target_mean = 5.5
n = 100
sd = 2  # Controls spread—adjust as needed

# Calculate `a` and `b` parameters for truncnorm (normalized bounds)
a = (min_val - target_mean) / sd
b = (max_val - target_mean) / sd

# Generate truncated normal random numbers
np.random.seed(123)
rand_nums = truncnorm.rvs(a, b, loc=target_mean, scale=sd, size=n)

# Verify results
print(f"Mean: {rand_nums.mean():.2f}")
print(f"Min: {rand_nums.min():.2f}")
print(f"Max: {rand_nums.max():.2f}")

Pro Tip: If the mean is off by a tiny bit, you can adjust the loc parameter slightly (e.g., loc=target_mean + 0.1) or tweak the sd to get closer to your target.

内容的提问来源于stack exchange，提问作者user37353