---
title: "Linear Regression Lab on the Boston and Carseats Data (stub)"
format: html
---

Stub of the Linear Regression Lab on the Boston and Carseats Data page: the exercises, without the solutions. Solutions: https://morrison-lab.github.io/lds/chapters/linear-regression-lab.html

#### Lab: fit a line to the Boston data

Download the Boston data from [`data/boston.csv`](https://morrison-lab.github.io/lds/chapters/data/boston.csv)
and save it as `data/boston.csv` next to your code.
It has $506$ Boston neighborhoods from the 1970 census.
The copy here is the one in the Python package `ISLP` and the R package `ISLR2`.
The response `medv` is the median home value, in thousands of dollars,
and `lstat` is the percentage of the population with lower socioeconomic status.

Fit `medv` on `lstat`.
Report each [coefficient](https://morrison-lab.github.io/lds/chapters/big-ideas.html#def-parameter) with its [standard error](https://morrison-lab.github.io/sds/estimation.html#def-SE).

::: {.callout-tip}

#### Hint

Read the data with the [data frame functions](https://morrison-lab.github.io/lds/chapters/programming-primer.html#exm-prog-pandas),
then [fit a model with a formula](https://morrison-lab.github.io/lds/chapters/programming-primer.html#exm-prog-statsmodels)
and read its [coefficients](https://morrison-lab.github.io/lds/chapters/big-ideas.html#def-parameter) and [standard errors](https://morrison-lab.github.io/sds/estimation.html#def-SE) from the results.

:::

```{python}
# your code here
```

#### Lab: fit a line to the Boston data (part 2)

Use the setup of Exercise 1.

[Predict](https://morrison-lab.github.io/lds/chapters/course-overview.html#def-prediction) `medv` for `lstat` values $5$, $10$ and $15$.

::: {.callout-tip}

#### Hint

A [fitted model](https://morrison-lab.github.io/lds/chapters/big-ideas.html#def-fitted-model) [predicts at new values](https://morrison-lab.github.io/lds/chapters/programming-primer.html#exm-prog-statsmodels) of the [feature](https://morrison-lab.github.io/lds/chapters/anatomy.html#def-feature)
when you pass them in a data frame with the same column name.

:::

```{python}
# your code here
```

#### Lab: fit a line to the Boston data (part 3)

Use the setup of Exercise 1.

Give the $95\%$ [confidence interval](https://morrison-lab.github.io/sds/inference.html#def-confidence-interval) for the [mean](https://morrison-lab.github.io/sds/exploratory-descriptive.html#def-sample-mean) [response](https://morrison-lab.github.io/lds/chapters/anatomy.html#def-output-variable) at each value.

::: {.callout-tip}

#### Hint

The [prediction object](https://morrison-lab.github.io/lds/chapters/programming-primer.html#exm-prog-statsmodels) of the [fit](https://morrison-lab.github.io/lds/chapters/big-ideas.html#def-fitted-model) holds the [confidence interval](https://morrison-lab.github.io/sds/inference.html#def-confidence-interval) of the [mean](https://morrison-lab.github.io/sds/exploratory-descriptive.html#def-sample-mean) [response](https://morrison-lab.github.io/lds/chapters/anatomy.html#def-output-variable) at each new value.

:::

```{python}
# your code here
```

#### Lab: fit a line to the Boston data (part 4)

Use the setup of Exercise 1.

Give the $95\%$ [prediction interval](https://morrison-lab.github.io/lds/chapters/linear-regression-inference.html#def-prediction-interval) at each value.

::: {.callout-tip}

#### Hint

The [prediction object](https://morrison-lab.github.io/lds/chapters/programming-primer.html#exm-prog-statsmodels) of the [fit](https://morrison-lab.github.io/lds/chapters/big-ideas.html#def-fitted-model) also holds the [prediction interval](https://morrison-lab.github.io/lds/chapters/linear-regression-inference.html#def-prediction-interval) at each new value.

:::

```{python}
# your code here
```

#### Lab: fit a line to the Boston data (part 5)

Use the setup of Exercise 1.

Why is the [prediction interval](https://morrison-lab.github.io/lds/chapters/linear-regression-inference.html#def-prediction-interval) wider?

::: {.callout-tip}

#### Hint

Compare what the [confidence interval](https://morrison-lab.github.io/sds/inference.html#def-confidence-interval) covers with what the [prediction interval](https://morrison-lab.github.io/lds/chapters/linear-regression-inference.html#def-prediction-interval) covers.

:::

```{python}
# your code here
```

#### Lab: fit a line to the Boston data (part 6)

Use the setup of Exercise 1.

Draw a scatter plot of `medv` against `lstat` with the [fitted](https://morrison-lab.github.io/lds/chapters/big-ideas.html#def-fitted-model) line.

::: {.callout-tip}

#### Hint

The [plotting functions](https://morrison-lab.github.io/lds/chapters/programming-primer.html#exm-prog-plots) draw the scatter plot,
and a [fitted line over the data](https://morrison-lab.github.io/lds/chapters/programming-primer.html#exm-prog-reporting) comes from [predictions](https://morrison-lab.github.io/lds/chapters/course-overview.html#def-prediction) on a grid of [feature](https://morrison-lab.github.io/lds/chapters/anatomy.html#def-feature) values.

:::

```{python}
# your code here
```

#### Lab: fit a line to the Boston data (part 7)

Use the setup of Exercise 1.

Draw the [residual plot](https://morrison-lab.github.io/lds/chapters/linear-regression-inference.html#def-residual-plot).

::: {.callout-tip}

#### Hint

The [residuals and fitted values](https://morrison-lab.github.io/lds/chapters/programming-primer.html#exm-prog-diagnostics) of a [fit](https://morrison-lab.github.io/lds/chapters/big-ideas.html#def-fitted-model) are the two axes of the [residual plot](https://morrison-lab.github.io/lds/chapters/linear-regression-inference.html#def-residual-plot).

:::

```{python}
# your code here
```

#### Lab: fit a line to the Boston data (part 8)

Use the setup of Exercise 1.

What do the plots say about a straight line for these data?

::: {.callout-tip}

#### Hint

A straight line suits the data
when the [residual plot](https://morrison-lab.github.io/lds/chapters/linear-regression-inference.html#def-residual-plot) shows no pattern.

:::

```{python}
# your code here
```

#### Lab: many features on the Boston data

Fit `medv` on `lstat` and `age`,
the percentage of owner-occupied homes built before 1940.
How does the `lstat` [coefficient](https://morrison-lab.github.io/lds/chapters/big-ideas.html#def-parameter) compare with that of Exercise 1?

::: {.callout-tip}

#### Hint

A [formula](https://morrison-lab.github.io/lds/chapters/programming-primer.html#exm-prog-statsmodels) names each [feature](https://morrison-lab.github.io/lds/chapters/anatomy.html#def-feature) after a `+`,
and the [coefficient](https://morrison-lab.github.io/lds/chapters/big-ideas.html#def-parameter) of `lstat` is in the [results](https://morrison-lab.github.io/lds/chapters/programming-primer.html#exm-prog-statsmodels).

:::

```{python}
# your code here
```

#### Lab: many features on the Boston data (part 2)

Use the setup of Exercise 9.

Fit `medv` on all $12$ other columns.

::: {.callout-tip}

#### Hint

A [formula](https://morrison-lab.github.io/lds/chapters/programming-primer.html#exm-prog-statsmodels) can list the other columns by name, or [build the formula as text from a list of the column names](https://morrison-lab.github.io/lds/chapters/programming-primer.html#exm-prog-statsmodels).

:::

```{python}
# your code here
```

#### Lab: many features on the Boston data (part 3)

Use the setup of Exercise 9.

Which [coefficients](https://morrison-lab.github.io/lds/chapters/big-ideas.html#def-parameter) have $p$-values above $0.05$?

::: {.callout-tip}

#### Hint

The [summary](https://morrison-lab.github.io/lds/chapters/programming-primer.html#exm-prog-statsmodels) of the [fit](https://morrison-lab.github.io/lds/chapters/big-ideas.html#def-fitted-model) holds a [$p$-value](https://morrison-lab.github.io/sds/inference.html#def-p-value) for each [coefficient](https://morrison-lab.github.io/lds/chapters/big-ideas.html#def-parameter),
and a [mask](https://morrison-lab.github.io/lds/chapters/programming-primer.html#exm-prog-pandas) keeps those above $0.05$.

:::

```{python}
# your code here
```

#### Lab: many features on the Boston data (part 4)

Use the setup of Exercise 9.

Compute the [variance inflation factor](https://morrison-lab.github.io/lds/chapters/linear-regression-inference.html#def-vif)
for each [feature](https://morrison-lab.github.io/lds/chapters/anatomy.html#def-feature) of the full [model](https://morrison-lab.github.io/lds/chapters/anatomy.html#def-model).

::: {.callout-tip}

#### Hint

The [variance inflation factor](https://morrison-lab.github.io/lds/chapters/linear-regression-inference.html#def-vif) of each [feature](https://morrison-lab.github.io/lds/chapters/anatomy.html#def-feature) [comes from the design matrix](https://morrison-lab.github.io/lds/chapters/programming-primer.html#exm-prog-statsmodels-glm) of the [fit](https://morrison-lab.github.io/lds/chapters/big-ideas.html#def-fitted-model).

:::

```{python}
# your code here
```

#### Lab: many features on the Boston data (part 5)

Use the setup of Exercise 9.

Which [feature](https://morrison-lab.github.io/lds/chapters/anatomy.html#def-feature) has the largest, and what is it?

::: {.callout-tip}

#### Hint

Look for the largest of the [variance inflation factors](https://morrison-lab.github.io/lds/chapters/linear-regression-inference.html#def-vif),
and name the [feature](https://morrison-lab.github.io/lds/chapters/anatomy.html#def-feature) that has it.

:::

```{python}
# your code here
```

#### Lab: an interaction on the Boston data

Fit `medv` on `lstat`, `age` and their product $\text{lstat} \times \text{age}$,
the [interaction](https://morrison-lab.github.io/lds/chapters/linear-regression-inference.html#def-interaction) of the two [features](https://morrison-lab.github.io/lds/chapters/anatomy.html#def-feature).
Report the coefficients with their $p$-values.

::: {.callout-tip}

#### Hint

In a [formula](https://morrison-lab.github.io/lds/chapters/programming-primer.html#exm-prog-statsmodels), `*` adds both [features](https://morrison-lab.github.io/lds/chapters/anatomy.html#def-feature) and their product.

:::

```{python}
# your code here
```

#### Lab: an interaction on the Boston data (part 2)

Use the setup of Exercise 14.

Is there evidence that the [slope](https://morrison-lab.github.io/mds/algebra-exponentials.html#def-slope) on `lstat` depends on `age`?

::: {.callout-tip}

#### Hint

Read the [$p$-value](https://morrison-lab.github.io/sds/inference.html#def-p-value) of the [interaction term](https://morrison-lab.github.io/lds/chapters/linear-regression-inference.html#def-interaction) in the [summary](https://morrison-lab.github.io/lds/chapters/programming-primer.html#exm-prog-statsmodels) of the [fit](https://morrison-lab.github.io/lds/chapters/big-ideas.html#def-fitted-model).

:::

```{python}
# your code here
```

#### Lab: a curve in `lstat`

Add $\text{lstat}^2$ to the [model](https://morrison-lab.github.io/lds/chapters/anatomy.html#def-model) of Exercise 9,
so the [features](https://morrison-lab.github.io/lds/chapters/anatomy.html#def-feature) are `lstat`, $\text{lstat}^2$ and `age`.
Report the [coefficients](https://morrison-lab.github.io/lds/chapters/big-ideas.html#def-parameter) with their $p$-values.

::: {.callout-tip}

#### Hint

Wrap the square in `I(...)` inside the [formula](https://morrison-lab.github.io/lds/chapters/programming-primer.html#exm-prog-statsmodels)
so that `**` is read as a power.
In R, write `^` instead of `**`.

:::

```{python}
# your code here
```

#### Lab: a curve in `lstat` (part 2)

Use the setup of Exercise 16.

Compare the two [models](https://morrison-lab.github.io/lds/chapters/anatomy.html#def-model) with the [partial $F$-test](https://morrison-lab.github.io/lds/chapters/linear-regression-inference.html#def-partial-f-test).

::: {.callout-tip}

#### Hint

The [`anova_lm` function](https://morrison-lab.github.io/lds/chapters/programming-primer.html#exm-prog-statsmodels-glm) compares two nested [models](https://morrison-lab.github.io/lds/chapters/anatomy.html#def-model) with an $F$ test.
In R, `anova()` does the same.

:::

```{python}
# your code here
```

#### Lab: a curve in `lstat` (part 3)

Use the setup of Exercise 16.

What does it say about the squared term?

::: {.callout-tip}

#### Hint

Read the [$p$-value](https://morrison-lab.github.io/sds/inference.html#def-p-value) of the $F$ test:
a small value says the larger [model](https://morrison-lab.github.io/lds/chapters/anatomy.html#def-model) [fit](https://morrison-lab.github.io/lds/chapters/big-ideas.html#def-fitted-model)s better than the smaller one.

:::

```{python}
# your code here
```

#### Lab: a curve in `lstat` (part 4)

Use the setup of Exercise 16.

Draw the [residual plot](https://morrison-lab.github.io/lds/chapters/linear-regression-inference.html#def-residual-plot) of the new [model](https://morrison-lab.github.io/lds/chapters/anatomy.html#def-model).

::: {.callout-tip}

#### Hint

The [residuals and fitted values](https://morrison-lab.github.io/lds/chapters/programming-primer.html#exm-prog-diagnostics) of the new [fit](https://morrison-lab.github.io/lds/chapters/big-ideas.html#def-fitted-model) are the two axes of its [residual plot](https://morrison-lab.github.io/lds/chapters/linear-regression-inference.html#def-residual-plot).

:::

```{python}
# your code here
```

#### Lab: a curve in `lstat` (part 5)

Use the setup of Exercise 16.

Compare it with the [residual plot](https://morrison-lab.github.io/lds/chapters/linear-regression-inference.html#def-residual-plot) of Exercise 1.

::: {.callout-tip}

#### Hint

Compare whether the [residual plot](https://morrison-lab.github.io/lds/chapters/linear-regression-inference.html#def-residual-plot) still shows the curve that the [residual plot](https://morrison-lab.github.io/lds/chapters/linear-regression-inference.html#def-residual-plot) of Exercise 1 showed.

:::

```{python}
# your code here
```

#### Lab: a category with three levels

Download the Carseats data from [`data/carseats.csv`](https://morrison-lab.github.io/lds/chapters/data/carseats.csv)
and save it as `data/carseats.csv` next to your code.
It is simulated data on sales of child car seats at $400$ stores,
from the book.
`Sales` is unit sales in thousands.
`ShelveLoc` is the quality of the shelf space for the car seats at each store:
`Bad`, `Medium` or `Good`.

Fit `Sales` on all the other columns,
plus the interactions $\text{Income} \times \text{Advertising}$
and $\text{Price} \times \text{Age}$.

::: {.callout-tip}

#### Hint

In a [formula](https://morrison-lab.github.io/lds/chapters/programming-primer.html#exm-prog-statsmodels), `Income:Advertising` adds only the product,
and a column of text is treated as a category automatically.

:::

```{python}
# your code here
```

#### Lab: a category with three levels (part 2)

Use the setup of Exercise 21.

Which indicator columns does the software build for `ShelveLoc`?

::: {.callout-tip}

#### Hint

The [category terms](https://morrison-lab.github.io/lds/chapters/programming-primer.html#exm-prog-statsmodels) of a formula name one indicator column for each level except the baseline;
read the names in the [summary](https://morrison-lab.github.io/lds/chapters/programming-primer.html#exm-prog-statsmodels).

:::

```{python}
# your code here
```

#### Lab: a category with three levels (part 3)

Use the setup of Exercise 21.

Which level has no column of its own?

::: {.callout-tip}

#### Hint

The baseline is the level whose indicator is missing from the [summary](https://morrison-lab.github.io/lds/chapters/programming-primer.html#exm-prog-statsmodels).

:::

```{python}
# your code here
```

#### Lab: a category with three levels (part 4)

Use the setup of Exercise 21.

Interpret the [coefficients](https://morrison-lab.github.io/lds/chapters/big-ideas.html#def-parameter) of the two `ShelveLoc` indicators.

::: {.callout-tip}

#### Hint

Each [coefficient](https://morrison-lab.github.io/lds/chapters/big-ideas.html#def-parameter) of an indicator is a difference from the baseline, with the other [features](https://morrison-lab.github.io/lds/chapters/anatomy.html#def-feature) held fixed.

:::

```{python}
# your code here
```
