Data Science Get Started

Setting Up Your Environment

Install the core stack with a single pip command, ideally inside a virtual environment so versions don't clash with other projects: pip install numpy pandas matplotlib scikit-learn scipy.

import numpy
import pandas
import matplotlib
import sklearn

print("numpy:", numpy.__version__)
print("pandas:", pandas.__version__)
print("matplotlib:", matplotlib.__version__)
print("scikit-learn:", sklearn.__version__)

• numpy gives you the ndarray, the fast fixed-type array almost everything else is built on.
• pandas gives you the DataFrame, a labeled table for loading and cleaning data.
• matplotlib.pyplot gives you plotting functions like scatter and hist.
• sklearn gives you ready-made model classes like LinearRegression with a fit/predict interface.

Your First End-to-End Script

A minimal data science script touches every stage of the pipeline at once: it holds data in arrays, computes a summary statistic, fits a simple model, and plots the result.

import numpy as np
import matplotlib.pyplot as plt
from scipy import stats

# a tiny "dataset": hours studied and exam score
hours = np.array([1, 2, 3, 4, 5, 6, 7, 8, 9, 10])
scores = np.array([50, 55, 65, 70, 63, 72, 80, 84, 90, 95])

print("average score:", np.mean(scores))

slope, intercept, r, p, std_err = stats.linregress(hours, scores)
print("best-fit line: score =", round(slope, 2), "* hours +", round(intercept, 2))

plt.scatter(hours, scores)
plt.plot(hours, slope * hours + intercept)
plt.xlabel("Hours studied")
plt.ylabel("Exam score")
plt.show()

How This Tutorial Is Organized

The chapters ahead move from describing a single column of numbers (distributions), to comparing two columns (scatter plots), to fitting lines and curves that predict one column from another (linear, polynomial, and multiple regression).

By the end, you'll be able to take a raw pair of columns and decide, with evidence rather than a guess, whether one predicts the other.