Hey there, I've put together the exact implementation you're asking for, with clear explanations for each step. Quick heads-up: the load_boston dataset has been deprecated in newer scikit-learn versions due to ethical issues around its original data, but I'll follow your specified requirements closely here.

1. Import Required Libraries #

First, we'll import all the libraries you mentioned:

import pandas as pd
import numpy as np
from sklearn.datasets import load_boston
import random

2. Load Dataset & Convert to DataFrame #

Next, we'll load the Boston housing data, convert the feature matrix into a pandas DataFrame, and add the target column MEDV (which represents the median value of owner-occupied homes in $1000s):

# Load the dataset
boston = load_boston()

# Convert features to DataFrame
df = pd.DataFrame(boston.data, columns=boston.feature_names)

# Add target column 'MEDV'
df['MEDV'] = boston.target

3. Generate 100 Random Samples (Each with 30 Observations) #

We'll set a random seed for reproducibility, then use random.sample to generate 100 separate samples, each containing 30 rows from our DataFrame:

# Set random seed for reproducibility
random.seed(1)

# Initialize a list to store all samples
samples_list = []

# Generate 100 samples
for _ in range(100):
    # Randomly select 30 row indices without replacement
    sample_indices = random.sample(list(df.index), k=30)
    # Extract the corresponding rows from the DataFrame
    sample = df.loc[sample_indices]
    # Add the sample to our list
    samples_list.append(sample)

Key Notes: #

• random.seed(1) ensures that every time you run this code, you'll get the exact same set of samples (critical for reproducibility).
• random.sample(list(df.index), k=30) picks 30 unique row indices from the DataFrame without repeating any rows in a single sample.
• Each item in samples_list is a separate pandas DataFrame with 30 rows and all original columns (including MEDV).

If you want to verify the output, you can run these quick checks:

# Check total number of samples generated
print(f"Total samples: {len(samples_list)}")
# Check the dimensions of the first sample
print(f"First sample shape: {samples_list[0].shape}")

内容的提问来源于stack exchange，提问作者akietta