GitHub

Original file line numberDiff line numberDiff line change

@@ -685,15 +685,15 @@ Numerical routines would in this case use the alternative form $R \hat \beta = Q

685685

## Exercises

686686
687687

```{exercise-start}

688-

:label: ex1

688+

:label: op_ex1

689689

```

690690
691691

Show that, for any linear subspace $S \subset \mathbb R^n$, $S \cap S^{\perp} = \{0\}$.

692692
693693

```{exercise-end}

694694

```

695695
696-

```{solution-start} ex1

696+

```{solution-start} op_ex1

697697

:class: dropdown

698698

```

699699

If $x \in S$ and $x \in S^\perp$, then we have in particular

@@ -703,7 +703,7 @@ that $\langle x, x \rangle = 0$, but then $x = 0$.

703703

```

704704
705705

```{exercise-start}

706-

:label: ex2

706+

:label: op_ex2

707707

```

708708

Let $P = X (X' X)^{-1} X'$ and let $M = I - P$. Show that

709709

$P$ and $M$ are both idempotent and symmetric. Can you give any

@@ -712,7 +712,7 @@ intuition as to why they should be idempotent?

712712

```{exercise-end}

713713

```

714714
715-

```{solution-start} ex2

715+

```{solution-start} op_ex2

716716

:class: dropdown

717717

```

718718

@@ -729,7 +729,7 @@ already the closest point in the subspace to itself).

729729
730730
731731

```{exercise-start}

732-

:label: ex3

732+

:label: op_ex3

733733

```

734734
735735

Using Gram-Schmidt orthogonalization, produce a linear projection of $y$ onto the column space of $X$ and verify this using the projection matrix $P := X (X' X)^{-1} X'$ and also using QR decomposition, where:

@@ -765,7 +765,7 @@ $$

765765

```

766766
767767
768-

```{solution-start} ex3

768+

```{solution-start} op_ex3

769769

:class: dropdown

770770

```

771771

Read the original on github.com ↗