The computational companion

Scan it.
Run it.
Change it.

Book QR targets, runnable scripts, and selected R/Python examples—organised around the printed code boxes.

57 code boxes28 R targets29 Python targets

chapter06 / multiple_regression

# R
fit <- lm(price ~ living_area + monthly_fee,
          data = apartments)

# Python
fit = smf.ols(
  "price ~ living_area + monthly_fee",
  data=apartments
).fit()

Same model · two ecosystems

Book-to-browser bridge

Every QR opens
inspectable source.

The printed QR codes point to small, box-level files that mirror the code in the book. Use those for quick reference; use the full chapter scripts to run an example end to end.

Chapter 02 · Box 01

Calculate correlation

Build Pearson’s coefficient from covariance and variances, then verify it with the built-in method.

df <- read.csv("apartment_price_data.csv")
xbar <- mean(df$living_area)
ybar <- mean(df$price)
s_xy <- sum((df$living_area-xbar) *
            (df$price-ybar)) / (n-1)
corr_builtin <- cor(df$living_area, df$price)

Python

df = pd.read_csv("apartment_price_data.csv")
xbar = df["living_area"].mean()
ybar = df["price"].mean()
s_xy = ((df["living_area"]-xbar) *
        (df["price"]-ybar)).sum() / (n-1)
corr_builtin = df["living_area"].corr(df["price"])

R source ↗

QR code for Chapter 2 correlation in Python

Python source ↗

Chapter 06 · Box 01

Fit multiple regression

Estimate the same conditional model and report heteroskedasticity-robust inference.

ols_model <- lm(
  price ~ living_area + monthly_fee +
          new_production, data = df)
coeftest(ols_model,
  vcov = vcovHC(ols_model, type = "HC0"))

Python

ols_model = smf.ols(
  "price ~ living_area + monthly_fee + "
  "new_production", data=df).fit()
robust = ols_model.get_robustcov_results(
  cov_type="HC0", use_t=True)

QR code for Chapter 6 multiple regression in R

R source ↗

QR code for Chapter 6 multiple regression in Python

Python source ↗

Chapter 13 · Box 01

Regression discontinuity

Construct the running variable, restrict the bandwidth, and estimate the first stage.

df$rv <- df$share_seats_left_last_election - .5
df$Z  <- ifelse(df$rv >= 0, 1, 0)
df$Zrv <- df$Z * df$rv
df_h <- df[abs(df$rv) <= .05, ]
rd_fs <- lm(left_coalition_last_term ~
            Z + rv + Zrv, data = df_h)

Python

df["rv"] = df["share_seats_left_last_election"] - .5
df["Z"] = (df["rv"] >= 0).astype(int)
df["Zrv"] = df["Z"] * df["rv"]
df_h = df.loc[df["rv"].abs() <= .05]
rd_fs = smf.ols(
  "left_coalition_last_term ~ Z + rv + Zrv",
  data=df_h).fit()