# Linear Regression¶

Due: Friday, Nov 10 at 4pm

In this assignment, you will fit linear regression models and implement a few simple variable selection algorithms. The assignment will give you experience with Numpy and more practice with using classes and functions to support code reuse.

You must work alone on this assignment.

At the heart of the assignment is a table, where each column is a variable and each row is a sample unit. As an example, in a health study, each sample unit might be a person, with variables like height, weight, sex, etc. In your analysis, you will build models that, with varying levels of accuracy, can predict the value of one of the variables as a function of the others.

Predictions are only possible if variables are related somehow. As an example, look at this plot of recorded crimes against logged complaint calls about garbage to 311.

Each point describes a sample unit, which in this example represents a geographical region of Chicago. Each region is associated with variables, such as the number of crimes or complaint calls during a fixed time frame. Given this plot, if you got the question of how many crimes you think were recorded for a region that had 150 complaint calls about garbage, you would follow the general trend and probably say something like 3000 recorded crimes. To formalize this prediction, we need a model for the data that relates a dependent variable (e.g., crimes) to a set of predictor variables (e.g., complaint calls). We will assume a linear dependence as follows:

$y_n = \beta_0 + \beta_1 x_{n1} + \dots + \beta_K x_{nK} + \varepsilon_n,$

or using vector notation

$y_n = \mathbf{x}_n^T \beta + \varepsilon_n.$

We have introduced the following notation:

$$n$$
The sample unit that we are considering out of $$N$$ total.
$$y_n$$
An observation of the dependent variable, e.g., the total number of crimes.
$$x_{nk}$$
An observation of a predictor variable $$k$$ for sample unit $$n,$$ e.g., the complaint calls about garbage. We assume a total of $$K$$ different predictors.
$$\beta = (\beta_0, \beta_1, \dots, \beta_K)^T$$
A column vector of the regression coefficients, where $$\beta_0$$ is the intercept and $$\beta_k$$ is the coefficient associated with the $$k\mathrm{th}$$ predictor. This vector describes the red line in the figure above. Note that a positive value of a coefficient suggests a positive correlation with the dependent variable. The same is true for a negative value and a negative correlation.
$$\varepsilon_n$$
Not all sample unit points lie on the red line, which means the equation $$y_n = \mathbf{x}_n^T \beta$$ is rarely satisfied. In our model, this discrepancy is assumed to be caused by random error. We assume that the random error is zero-mean and drawn from a normal distribution of a fixed variance. We want as much as possible to be explained deterministically, which means we will favor values of $$\beta$$ that result in small errors.
$$\mathbf{x}_n = (1, x_{n1}, \dots, x_{nK})^T$$
A column vector representation of all the predictors for a given sample unit. As a convenience, a 1 has been prepended to the vector to incorporate the intercept $$\beta_0$$ into the vector notation.

When making predictions, what we are really doing is asking what is the weighted average value for a given set of predictors. Since we have assumed the errors have zero mean, to make a prediction, we just remove the error term as:

$\hat y = f(\mathbf{x}) = \mathbf{x}^T \beta.$

This equation can be written for several sample units at the same time as:

$\mathbf{\hat y} = \mathbf{X} \beta,$

where $$\mathbf{\hat y}$$ is a column vector of predictions and $$\mathbf{X}$$ is a $$N \times K$$ matrix where each row is one sample unit.

## Model fitting¶

There are many possible candidates for $$\beta,$$ some that fit the data better than others. Finding the best value of $$\beta$$ is referred to as fitting the model. To simplify your job, we have provided code for fitting a linear model. The function linear_regression finds a value $$\beta$$ that minimizes $$\mathbf{\hat y} - \mathbf{y}$$ in a least squared sense. That is, it finds values for $$\beta$$ such that the predicted values $$\mathbf{\hat y} = \mathbf{X} \beta$$ are as close to the observed variables as possible (in a statistically motivated way using maximum likelihood). This function accepts a two-dimensional Numpy array of floats containing the predictor variables (X) and a one-dimensional Numpy array of floats (y) containing the dependent variable as input and returns $$\beta.$$ Note that X does not correspond exactly with $$\mathbf{X},$$ since it should not have the column of ones (we will add that for you).

We have also provided a function, apply_beta(beta, X) for making predictions using the model. It takes beta, the array generated by our linear regression function, and X, a two-dimensional Numpy array of floats that contains values for the predictor variables (again, without the added column of 1’s), applies the function described by beta to each row in X, and returns a one-dimensional Numpy array of the resulting values.

Here are sample calls to linear_regression and apply_beta:

>>> import numpy as np
>>> from model import *
>>> X = np.array([[5,2], [3,2], [6,2.1], [7,3]]) # predictors
>>> y = np.array([5,2,6,6]) # dependent
>>> beta = linear_regression(X, y)
>>> beta
array([ 1.20104895,  1.41083916, -1.6958042 ]) # f(x) = 1.2 + 1.4*x1 - 1.7*x2
>>> apply_beta(beta, X)
array([ 4.86363636,  2.04195804,  6.1048951 ,  5.98951049]) # predictions


The input to both of these functions require X to be two-dimensional, even in the case of K = 1. Numpy makes a distinction between one-dimensional and two-dimensional arrays, even if they contain the same information. For instance, if we want to extract the second column of X as a one-dimensional array, we do the following:

>>> X
array([[ 5. ,  2. ],
[ 3. ,  2. ],
[ 6. ,  2.1],
[ 7. ,  3. ]])
>>> X[:,1]
array([ 2. ,  2. ,  2.1,  3. ])
>>> X[:,1].shape
(4,)


The resulting shape will not be accepted by linear_regression and apply_beta. To retrieve a 2D column subset of X, you can use a list of integers as the index. This mechanism keeps X two-dimensional, even if the list of indices has only one value:

>>> X[:,[1]]
array([[ 2. ],
[ 2. ],
[ 2.1],
[ 3. ]])
>>> X[:,[1]].shape
(4, 1)


In general, you can specify a slice containing a specific subset of columns as a list. For example, let Z be:

>>> Z = np.array([[1, 2, 3, 4], [11, 12, 13, 14], [21, 22, 23, 24]])
>>> Z
array([[ 1,  2,  3,  4],
[11, 12, 13, 14],
[21, 22, 23, 24]])


Evaluating the expression Z[:, [0,2,3]] will yield a new 2D array with columns 0, 2, and 3 from the array Z:

>>> Z[:, [0, 2, 3]]
array([[ 1,  3,  4],
[11, 13, 14],
[21, 23, 24]])


or more generally, we can specify any expression for the slice that yields a list:

>>> l = [0, 2, 3]

>>> X[:, l]
array([[ 1,  3,  4],
[11, 13, 14],
[21, 23, 24]])


## Data¶

In this assignment you will write code that can be used on many different datasets. We have provided a couple of sample datasets for testing purposes.

City: Predicting crime rate from the number of complaint calls to 311. Predicting a stock index from a set of stocks.

For this assignment, a dataset is stored in a directory that contains two files: a CSV (comma-separated values) file and a JSON file. The CSV file contains a table where each column corresponds to a variable and each row corresponds to a sample unit. The first row contains the column names. The JSON file contains a few parameters:

• name: the name of the dataset,
• predictor_vars: the column indices of the predictor variables,
• dependent_var: the column index of the dependent variable,
• training_fraction: the fraction of the data that should be used to train models, and
• seed: a random number generator seed.

The last two parameters will be used to help split the data into two sets: one that you will use to fit or train models and one that you will use to evaluate how well different models predict outcomes on new-to-the-model data. We describe this process below.

## Submission requirements¶

You will do six tasks in this assignment. Tasks 1-5 list specific submission requirements. Code to generate the required output text should be added to the file output.py. When we test your code, we will run the commands:

$python3 output.py data/city$ python3 output.py data/stock


then manually inspect the output to the terminal. Submissions that do not include the necessary code in output.py will lose substantial credit.

All output must be labeled with both the name of the data set and the task number, and it should be output using print.

Most tasks require you to include one or more models in the output. These models should be output in the following form: the name of the dependent variable followed by a tilde (~) and the regression equation with the constant first. For example, here’s the expected output for the model shown above for predicting crime from calls about garbage:

CRIME_TOTALS ~ 144.003954 + 17.365791 * GARBAGE


## Getting started¶

We have seeded your repository with a directory for this assignment. To pick it up, change to your cmsc12100-aut-17-username directory (where the string username should be replaced with your username) and then run the command: git pull upstream master. You should also run git pull to make sure your local copy of your repository is in sync with the server.

The pa5 directory contains the following files:

dataset.py: Python file that contains code for reading data from JSON and CSV files. You will complete the DataSet class in this file. Python file that contains code for computing linear regressions and applying them to data. You should add your Model class to this file along with your general purpose model-selection functions. Python file where you will put your code to compute the required output. This file should rely heavily on the general functions that you add to model.py.

The pa5 directory also contains a data directory which, in turn, contains two sub-directories: city and stock.

We will be using the scikit-learn machine learning library in this assignment. It is installed on the department’s Linux machines, but not on the class VM. To install it on your VM, run:

sudo pip3 install sklearn


You will be asked for a password. Use the student account password: uccs.

Representing Datasets

Recall that each dataset has two files: a CSV file with the data to be used for training and testing models and a JSON file with the parameters for the assignment. In the file dataset.py, we have provided two functions: one to read the sample data from the CSV file and one to read the parameters from the JSON file. In this file, we have also defined a skeleton for a class, named DataSet, for representing a dataset.

Your first task is to complete the DataSet constructor method, which takes the name of the directory holding the dataset as its only parameter. In addition to saving the values necessary for the computation from the JSON and CSV files (dataset name, predictor variable indices, dependent variable index, and the column labels), your constructor should also split the sample data into two Numpy arrays: one, called the training data, that you will use to construct models, and another, called the testing data, that you will use to evaluate how well different models predict new-to-the-model data.

Why not use all the data to fit the models? It is easy to fit a model that does a good job of predicting the dependent variable for the sample units that were used to train it, but does a terrible job of predicting the dependent variable for other data. How do we determine how well a model does with new-to-it data? We can hold out some of the data and use the model to predict the values of the dependent variable for the held-out sample units. Since we know the actual value of the dependent variable for this held-out data, we can use it to evaluate the effectiveness of the model on previously-unseen data.

We will use the train_test_split function from the Scikit-learn model selection library to make the split. This method takes a Numpy array and splits it into two arrays: one to use for training models and another to use for testing them. It has two keyword parameters that can be used to control the size of each array: train_size and test_size. We’ll set train_size using the training_fraction parameter from the parameters file and we will set the test_size parameter to None. Using None for the test_size parameter indicates that the rows not chosen for the training set should be included in the test set, which is the default behavior. Note that you could omit this parameter, but the function generates an annoying warning message if you do.

The train_test_split method decides randomly which rows to include in which set. To support repeatability and simplify testing, it includes a random_state parameter that can be used to set the seed for the underlying random process. Starting from a given seed, the “random” (actually pseudo-random) function will generate the same “random” numbers in the same order each time, causing the same rows to be assigned to the same sets each time you run your program. We’ll set this parameter using the seed parameter from the JSON file.

Your implementation of the DataSet constructor should be quite simple. Ours is less than 10 lines of code. The heavy lifting is done in the IO functions we provided, as well as the train_test_split method from Scikit-learn.

Representing models

In this assignment, you will be creating many different models for a given dataset using different subsets of the dataset’s predictor variables. You are required to add a class to model.py for representing such models.

Each model will be constructed from a dataset and a subset of the dataset’s predictor variables. We encourage you to think carefully about what attributes and methods to include in your class. Don’t be discouraged if you find yourself reworking the design of your class several times as you work through the tasks for this assignment.

Other than the requirement to add classes for representing the dataset and the models, you have free rein over the structure of your code. Think carefully about what functions to build, since these choices along with the design of your model class will largely determine the quality of your code. We suggest that you create functions that are as general as possible and break them up into several smaller function if they get too big.

## Task 1: Model evaluation using the Coefficient of Determination (R2)¶

Blindly regressing on all available predictors is bad practice and can lead to over-fit models. Over-fitting happens when your model picks up on random structure in the training data that does not represent the underlying phenomenon. This situation is made even worse if some of your predictors are highly correlated (see multicollinearity), since it can ruin the interpretation of your coefficients. As an example, consider a study that looks at the effects of smoking and drinking on developing lung cancer. Smoking and drinking are highly correlated, so the regression can arbitrarily pick up either as the explanation for high prevalence of lung cancer. As a result, drinking could be assigned a high coefficient while smoking a low one, leading to the questionable conclusion that drinking causes lung cancer while smoking does not.

This example illustrates the importance of variable selection. In order to compare two different models (subsets of predictors) against each other, we need a measurement of their goodness of fit, i.e. how well they explain the data. A popular choice is the coefficient of determination, R2, which first considers the variance of the dependent variable, $$\mathrm{Var}(y)$$:

In this example we revisit the regression of the total number of crimes, y, on the number of calls about garbage to 311, x. To the right, a histogram over y is shown and its variance is calculated. The variance is a measure of how spread out the data is. It is estimated from sample units as:

$\mathrm{Var}(y) = \frac{1}{N}{\sum_{n=1}^N} (y_n - \bar y)^2,$

where $$\bar y$$ denotes the mean of all y’s.

We now subtract the red line from each point. These new points are called the residuals and represent the deviations from the expected values under our model. If the variance of the residuals, $$\mathrm{Var}(y - \hat y),$$ is zero, it means that all residuals had to be zero and thus all sample units lie perfectly on the fitted line. On the other hand, if the variance of the residuals is the same as the variance of y, it means that the model did not explain anything away and we can consider the model useless. These two extremes represent R2 values of 1 and 0, respectively. All R2 values in between these two extremes will be cases where the variance was reduced some, but not entirely. Our example is one such case:

As we have hinted, the coefficient of determination, R2, measures how much the variance was reduced by subtracting the model. More specifically, it calculates this reduction as a percentage of the original variance:

$R^2 = \frac{\mathrm{Var}(y) - \mathrm{Var}(y - \hat y)}{\mathrm{Var}(y)} = 1 - \frac{\mathrm{Var}(y - \hat y)}{\mathrm{Var}(y)},$

which finally becomes:

$R^2 = 1 - \frac{\sum_{n=1}^N (y_n - \hat y_n)^2}{\sum_{n=1}^N (y_n - \bar y)^2}. \quad (1)$

In these steps we omitted the normalization constants N since they cancel each other out. We also did not subtract the mean when calculating the variance of the residuals, since it can be shown mathematically that if the model has been fit by least squares, the sum of the residuals is always zero.

There are two properties of R2 that are good to keep in mind when checking the correctness of your calculations:

1. 0 ≤ R2 ≤ 1
2. If model A contains a superset of predictors of model B, then model A cannot have a lower R2 value than B.

These properties are only true if R2 is computed using the same data that was used to train the model parameters. If you calculate R2 on a held-out testing set, the R2 value could decrease with more predictors (due to over-fitting) and R2 is no longer guaranteed to be greater than or equal to zero. The intuition behind R2 would also need to be reconsidered, but since the final formula in (1) remains the same we will omit the details here.

A good model should explain as much of the variance as possible, so we will be favoring higher R2 values. However, since using all predictors gives the highest R2, we must balance this goal with a desire to use few predictors. In general it is important to be cautious and not blindly interpret R2; take a look at these two generated examples:

To the left, we see that the fitted model is close to the one we used to generate the data. However, the variance of the error is still high, so the R2 is rather low. To the right is a model that clearly is not right for the data, but still manages to record a high R2 value. This example also shows that that linear regression is not always the right model choice. However, in many cases, a simple transformation of the predictors or the dependent variable can resolve this problem, making the linear regression paradigm still relevant.

In this task, you will construct P univariate (i.e., one variable) models of the dataset’s dependent variable, one for each of the possible predictor variables, and compute their associated R2 values. Note that when we ask for the R2 value for a model, we mean the value obtained by computing R2 using the model’s $$\beta$$ on the data that was used to train it.

Submission requirement 1a

• Output a table of P models, one per possible predictor variable, and their associated R2 values.

Note

You may not use the Numpy var method to compute R2 for two reasons: (1) our computation uses the biased variance whereas the var method computes the unbiased variance and will lead you to generate the wrong answers and (2) we want you to get practice with using operations on Numpy arrays.

From this table, you will see that none of the predictor variables individually are particularly good predictors for the dependent variable. In subsequent tasks, we will construct better models using multiple variables. For example, we could predict crime using not only complaints about garbage, but graffiti as well. In this example, we want to find $$\beta = (\beta_0, \beta_\mathrm{garbage}, \beta_\mathrm{graffiti})$$ in the equation

$f(\mathbf{x}) = \beta_0 + \beta_\mathrm{garbage} x_\mathrm{garbage} + \beta_\mathrm{graffiti} x_\mathrm{graffiti},$

where $$f(\mathbf{x})$$ is a prediction of the number of crimes given the number of complaint calls about garbage $$(x_\mathrm{garbage})$$ and graffiti $$(x_\mathrm{graffiti}).$$

In terms of the R2 calculation, this new equation only changes how each prediction $$\hat y_n$$ is calculated. At this point, you should make sure that your code to calculate a model’s R2 value is general enough to handle models of with multiple predictor variables (i.e. multivariate models).

Submission requirement 1b

• Construct a single model that uses all of the dataset’s predictor variables to predict the dependent variable. Output the model along with its R2 value. According to the second property of R2, this value should be the largest R2 value you will see for the training data.

## Task 2: Building and selecting bivariate models¶

In Task 1, we found that none of our single variable models were particularly accurate (they all had low R2 values compared to when using all predictors). In this and future tasks we’re going to build more complex models that use multiple variables as predictors, without using all of them and risking over-fitting.

For this task, you will test all possible bivariate models (K = 2) and determine the one with the highest R2 value. We suggest that you use two nested for-loops to iterate over all combinations of two predictors, calculate the R2 value for each combination and keep track of one with the highest R2 value.

How does your result compare to the single-variable models in Task 1? Which pair of variables perform best together? Does this result make sense given the results from Task 1?

Submission requirement

• Output the bivariate model with the highest R2 value along with its R2 value.

## Task 3: Building models of arbitrary complexity¶

How do we determine how many and which variables will generate the best model? It turns out that this question is unsolved and is of interest to both computer scientists and statisticians. To find the best model of three variables we could repeat the process in Task 2 with three nested for-loops instead of two. For K variables we would have to use K nested for-loops. Nesting for-loops quickly becomes computationally expensive and infeasible for even the most powerful computers. Keep in mind that models in genetics research easily reach into the thousands of variables. How then do we find optimal models of several variables?

We’ll approach this problem by using heuristics, which will give us an approximate (as opposed to an exact) result. We are going to split this task into two pieces: (1) determine the best K variables to use for a fixed K and (2) determine the best value for K.

How do we determine which K variables will yield the best model for a fixed K? As noted, the naive approach, test all possible combinations as in Task 2, is intractable for large K. An alternative simpler approach is to choose the K variables with the K highest R2 values in the table from Task 1. This approach does not work well if your predictors are correlated. As you’ve already seen from the first two tasks, neither of the top two individual predictors for crime (GARBAGE and STREET_LIGHTS) are included in the best bivariate model (CRIME_TOTALS ~ -36.151629 + 3.300180 * POT_HOLES + 7.129337 * ABANDONED_BUILDINGS).

Instead, you’ll implement a heuristic known as Backward elimination, which is an example of a greedy algorithm. Backward elimination starts with a set that contains all potential predictor variables and then repeatedly eliminates variables until the set contains K variables. At each step, the algorithm requires you to identify the variable in the model that when subtracted, yields the model with the best R2 value and eliminate it from the set.

Submission requirement

• Let P be the number of predictor variables in your data set. Output a table that lists K, the best K-variable model chosen by the backward elimination algorithm, and R2 for this model for $$1 \leq K \leq P$$.

## Task 4: Selecting the best K¶

In Task 3, you computed a list of the best K-variable models for $$1\leq K \leq P$$ using backward elimination. How do we choose among these models? The calculation of R2 does not take into account the number of variables in a model and so, while it is a valid way to choose among different models, each with K variables, it is not a good way to choose among different values for K. Adjusted R2, which does take the number of predictor variables in the models into account, is a reasonable alternative.

Essentially, adjusted R2 replaces the biased variances used in R2 with their unbiased counterparts.) Here’s a formulation for it that highlights the adjustment, rather than the change in the definition of variance:

$adj\_R^2 = R^2 - (1 - R^2) * \frac{K}{(N-K-1)}$

where $$N$$ is the sample size (i.e. number of rows in the training data) and $$K$$ is the number of predictor variables in the model.

Submission requirement

• Output the model computed using the backward elimination heuristic that has the largest adjusted R2 along with both its R2 and adjusted R2 values.

## Task 5: Training vs. test data¶

Until now, you have evaluated a model using the data that was used to train it. The resulting model may be quite good for that particular dataset, but it may not be particularly good at predicting novel data. This is the problem that we have been referring to as over-fitting.

Up to this point, we have only computed R2 values for the training data that was used to construct the model. It is also valid to compute an R2 value for a model when applied to other data. For this task, you will use the *testing data* to evaluate the model identified in Task 4. You will receive *zero credit* for this task if you train a new model using the testing data.

Submission requirements

• Output the model chosen for Task 4 along with its R2 value computed using the training data and its R2 value computed using the testing data.

## Submission requirement round-up¶

Here is a round-up of all the submission requirements. Your implementation of output.py must produce these results when called from the Linux command line with a path to the dataset.

Recall that you should include the data set name and task numbers in your output. See the City dataset for sample output.

• Output a table of P models, one per possible predictor variable, and their associated R2 values.
• Construct a single model that uses all of the dataset’s predictor variables to predict the dependent variable. Output the model along with its R2 value. According to the second property of R2, this value should be the largest R2 value you will see for the training data.

• Output the bivariate model with the highest R2 value along with its R2 value.

• Let P be the number of predictor variables in your data set. Output a table that lists K, the best K-variable model chosen by the backward elimination algorithm, and R2 for this model for $$1 \leq K \leq P$$.

• Output the model computed using the backward elimination heuristic that has the largest adjusted R2 along with both its R2 and adjusted R2 values.

• Output the model chosen for Task 4 along with its R2 value computed using the training data and its R2 value computed using the testing data.

## Submission¶

To submit your assignment, make sure that you have:

• registered for the assignment using chisubmit,
• added, committed, and pushed your code to the git server, and
• run the chisubmit submission command.
chisubmit student assignment register pa5

git commit -m "final version of pa5 ready for submission"
git push

chisubmit student assignment submit pa5


Acknowledgments: This assignment has been worked on by many people over the years, including Matthew Rocklin and Gustav Larsson.