Exercise 1: Why Raw Coefficients Can't Be Compared, and What Standardization Fixes — Possible Solution
====================================================================

WHY THE RAW UNITS ARE INCOMPARABLE, PER THIS CHAPTER
------------------------------
Per this chapter's own warn-box, "mileage is measured in tens of
thousands; year is a four-digit number changing by at most a handful of
units across the whole dataset. Comparing the two raw coefficients
directly would be comparing 'dollars per mile' against 'dollars per
year' — two genuinely different units, not a fair size comparison."

WHY THIS MAKES A DIRECT SIZE COMPARISON MEANINGLESS
------------------------------
A coefficient's raw size reflects both how strongly a feature actually
matters AND the units that feature happens to be measured in — a
coefficient of, say, -0.05 for mileage isn't inherently "small" just
because the number itself is small; it's applied to values in the tens
of thousands, so a modest-looking coefficient can still represent a
large real effect on price across the actual range of mileage values
in the data. A coefficient for year could look numerically large in
comparison purely because year's own values span a much smaller numeric
range (a handful of units) — a difference driven entirely by the
choice of units, not by which feature genuinely matters more for
predicting price.

WHAT STANDARDIZATION ACTUALLY DOES
------------------------------
Per this chapter's own code, StandardScaler rescales each feature onto
the same typical range before fitting — putting mileage and year onto
directly comparable numeric scales, regardless of their original,
wildly different units. Once both features are standardized, a
one-unit change in "standardized mileage" and a one-unit change in
"standardized year" represent the same kind of thing (a comparable
amount of typical variation within each feature), which is exactly
what makes comparing the resulting coefficients' own sizes a fair,
apples-to-apples comparison rather than an apples-to-oranges one.

WHY THIS MATTERS FOR THIS CHAPTER'S OWN CENTRAL FINDING
------------------------------
Without standardization first, any claim about mileage "mattering more"
than year based on raw coefficient size would be unreliable — the
apparent difference could simply be an artifact of the two features'
own different units, rather than a genuine reflection of which one
actually predicts price more strongly. Standardizing first is precisely
what allows this chapter's own finding-box conclusion to be trusted as
a real answer to ds1-9's own hypothesis, rather than a coincidence of
unit choice.

WHY THIS WORKS AS AN ANSWER
------------------------------
It explains precisely why mileage and year's own raw units make their
coefficients structurally incomparable, and explains what
standardization changes (putting both features on the same scale)
that makes a fair size comparison — and this chapter's own central
finding — actually valid.