fsrftest

Univariate feature ranking for regression using F-tests

Description

idx = fsrftest(Tbl,ResponseVarName) ranks features (predictors) using F-tests. The table Tbl contains predictor variables and a response variable, and ResponseVarName is the name of the response variable in Tbl. The function returns idx, which contains the indices of predictors ordered by predictor importance, meaning idx(1) is the index of the most important predictor. You can use idx to select important predictors for regression problems.

idx = fsrftest(Tbl,formula) specifies a response variable and predictor variables to consider among the variables in Tbl by using formula.

example

idx = fsrftest(Tbl,Y) ranks predictors in Tbl using the response variable Y.

example

idx = fsrftest(X,Y) ranks predictors in X using the response variable Y.

idx = fsrftest(___,Name,Value) specifies additional options using one or more name-value pair arguments in addition to any of the input argument combinations in the previous syntaxes. For example, you can specify categorical predictors and observation weights.

example

[idx,scores] = fsrftest(___) also returns the predictor scores scores. A large score value indicates that the corresponding predictor is important.

Examples

collapse all

Rank predictors in a numeric matrix and create a bar plot of predictor importance scores.

Load the sample data.

load robotarm.mat

The robotarm data set contains 7168 training observations (Xtrain and ytrain) and 1024 test observations (Xtest and ytest) with 32 features [1][2].

Rank the predictors using the training observations.

[idx,scores] = fsrftest(Xtrain,ytrain);

The values in scores are the negative logs of the p-values. If a p-value is smaller than eps(0), then the corresponding score value is Inf. Before creating a bar plot, determine whether scores includes Inf values.

find(isinf(scores))
ans =

  1x0 empty double row vector

scores does not include Inf values. If scores includes Inf values, you can replace Inf by a large numeric number before creating a bar plot for visualization purposes. For details, see Rank Predictors in Table.

Create a bar plot of the predictor importance scores.

bar(scores(idx))
xlabel('Predictor rank')
ylabel('Predictor importance score')

Select the top five most important predictors. Find the columns of these predictors in Xtrain.

idx(1:5)
ans = 1×5

    30    24    10     4     5

The 30th column of Xtrain is the most important predictor of ytrain.

Rank predictors in a table and create a bar plot of predictor importance scores.

If your data is in a table and fsrftest ranks a subset of the variables in the table, then the function indexes the variables using only the subset. Therefore, a good practice is to move the predictors that you do not want to rank to the end of the table. Move the response variable and observation weight vector as well. Then, the indexes of the output arguments are consistent with the indexes of the table. You can move variables in a table using the movevars function.

This example uses the Abalone data [3][4] from the UCI Machine Learning Repository [5]. Download the data and save it in your current folder with the name 'abalone.data'.

Store the data in a table.

tbl = readtable('abalone.data','Filetype','text','ReadVariableNames',false);
tbl.Properties.VariableNames = {'Sex','Length','Diameter','Height', ...
    'WWeight','SWeight','VWeight','ShWeight','NoShellRings'};

Preview the first few rows of the table.

head(tbl)
ans=8×9 table
     Sex     Length    Diameter    Height    WWeight    SWeight    VWeight    ShWeight    NoShellRings
    _____    ______    ________    ______    _______    _______    _______    ________    ____________

    {'M'}    0.455      0.365      0.095      0.514     0.2245      0.101       0.15           15     
    {'M'}     0.35      0.265       0.09     0.2255     0.0995     0.0485       0.07            7     
    {'F'}     0.53       0.42      0.135      0.677     0.2565     0.1415       0.21            9     
    {'M'}     0.44      0.365      0.125      0.516     0.2155      0.114      0.155           10     
    {'I'}     0.33      0.255       0.08      0.205     0.0895     0.0395      0.055            7     
    {'I'}    0.425        0.3      0.095     0.3515      0.141     0.0775       0.12            8     
    {'F'}     0.53      0.415       0.15     0.7775      0.237     0.1415       0.33           20     
    {'F'}    0.545      0.425      0.125      0.768      0.294     0.1495       0.26           16     

The last variable in the table is a response variable.

Rank the predictors in tbl. Specify the last column NoShellRings as a response variable.

[idx,scores] = fsrftest(tbl,'NoShellRings')
idx = 1×8

     3     4     5     7     8     2     6     1

scores = 1×8

  447.6891  736.9619       Inf       Inf       Inf  604.6692       Inf       Inf

The values in scores are the negative logs of the p-values. If a p-value is smaller than eps(0), then the corresponding score value is Inf. Before creating a bar plot, determine whether scores includes Inf values.

idxInf = find(isinf(scores))
idxInf = 1×5

     3     4     5     7     8

scores includes five Inf values.

Create a bar plot of predictor importance scores. Use the predictor names for the x-axis tick labels.

bar(scores(idx))
xlabel('Predictor rank')
ylabel('Predictor importance score')
xticklabels(strrep(tbl.Properties.VariableNames(idx),'_','\_'))
xtickangle(45)

The bar function does not plot any bars for the Inf values. For the Inf values, plot bars that have the same length as the largest finite score.

hold on
bar(scores(idx(length(idxInf)+1))*ones(length(idxInf),1))
legend('Finite Scores','Inf Scores')
hold off

The bar graph displays finite scores and Inf scores using different colors.

Input Arguments

collapse all

Sample data, specified as a table. Multicolumn variables and cell arrays other than cell arrays of character vectors are not allowed.

Each row of Tbl corresponds to one observation, and each column corresponds to one predictor variable. Optionally, Tbl can contain additional columns for a response variable and observation weights.

A response variable can be a categorical, character, or string array, logical or numeric vector, or cell array of character vectors. If the response variable is a character array, then each element of the response variable must correspond to one row of the array.

  • If Tbl contains the response variable, and you want to use all remaining variables in Tbl as predictors, then specify the response variable by using ResponseVarName. If Tbl also contains the observation weights, then you can specify the weights by using Weights.

  • If Tbl contains the response variable, and you want to use only a subset of the remaining variables in Tbl as predictors, then specify the subset of variables by using formula.

  • If Tbl does not contain the response variable, then specify a response variable by using Y. The response variable and Tbl must have the same number of rows.

If fsrftest uses a subset of variables in Tbl as predictors, then the function indexes the predictors using only the subset. The values in the 'CategoricalPredictors' name-value pair argument and the output argument idx do not count the predictors that the function does not rank.

fsrftest considers NaN, '' (empty character vector), "" (empty string), <missing>, and <undefined> values in Tbl for a response variable to be missing values. fsrftest does not use observations with missing values for a response variable.

Data Types: table

Response variable name, specified as a character vector or string scalar containing the name of a variable in Tbl.

For example, if a response variable is the column Y of Tbl (Tbl.Y), then specify ResponseVarName as 'Y'.

Data Types: char | string

Explanatory model of the response variable and a subset of the predictor variables, specified as a character vector or string scalar in the form 'Y ~ X1 + X2 + X3'. In this form, Y represents the response variable, and X1, X2, and X3 represent the predictor variables.

To specify a subset of variables in Tbl as predictors, use a formula. If you specify a formula, then fsrftest does not rank any variables in Tbl that do not appear in formula.

The variable names in the formula must be both variable names in Tbl (Tbl.Properties.VariableNames) and valid MATLAB® identifiers. For details, see Tips.

Data Types: char | string

Response variable, specified as a numeric, categorical, or logical vector, a character or string array, or a cell array of character vectors. Each row of Y represents the labels of the corresponding row of X.

fsrftest considers NaN, '' (empty character vector), "" (empty string), <missing>, and <undefined> values in Y to be missing values. fsrftest does not use observations with missing values for Y.

Data Types: single | double | categorical | logical | char | string | cell

Predictor data, specified as a numeric matrix. Each row of X corresponds to one observation, and each column corresponds to one predictor variable.

Data Types: single | double

Name-Value Pair Arguments

Specify optional comma-separated pairs of Name,Value arguments. Name is the argument name and Value is the corresponding value. Name must appear inside quotes. You can specify several name and value pair arguments in any order as Name1,Value1,...,NameN,ValueN.

Example: 'NumBins',20,'UseMissing',true sets the number of bins as 20 and specifies to use missing values in predictors for ranking.

List of categorical predictors, specified as the comma-separated pair consisting of 'CategoricalPredictors' and one of the values in this table.

ValueDescription
Vector of positive integersEach entry in the vector is an index value corresponding to the column of the predictor data (X or Tbl) that contains a categorical variable.
Logical vectorA true entry means that the corresponding column of predictor data (X or Tbl) is a categorical variable.
Character matrixEach row of the matrix is the name of a predictor variable. The names must match the names in Tbl. Pad the names with extra blanks so each row of the character matrix has the same length.
String array or cell array of character vectorsEach element in the array is the name of a predictor variable. The names must match the names in Tbl.
'all'All predictors are categorical.

By default, if the predictor data is in a table (Tbl), fsrftest assumes that a variable is categorical if it is a logical vector, unordered categorical vector, character array, string array, or cell array of character vectors. If the predictor data is a matrix (X), fsrftest assumes that all predictors are continuous. To identify any other predictors as categorical predictors, specify them by using the 'CategoricalPredictors' name-value pair argument.

If fsrftest uses a subset of variables in Tbl as predictors, then the function indexes the predictors using only the subset. The 'CategoricalPredictors' values do not count the predictors that the function does not rank.

Example: 'CategoricalPredictors','all'

Data Types: single | double | logical | char | string | cell

Number of bins for binning continuous predictors, specified as the comma-separated pair consisting of 'NumBins' and a positive integer scalar.

Example: 'NumBins',50

Data Types: single | double

Indicator for whether to use or discard missing values in predictors, specified as the comma-separated pair consisting of 'UseMissing' and either true to use or false to discard missing values in predictors for ranking.

fsrftest considers NaN, '' (empty character vector), "" (empty string), <missing>, and <undefined> values to be missing values.

If you specify 'UseMissing',true, then fsrftest uses missing values for ranking. For a categorical variable, fsrftest treats missing values as an extra category. For a continuous variable, fsrftest places NaN values in a separate bin for binning.

If you specify 'UseMissing',false, then fsrftest does not use missing values for ranking. Because fsrftest computes importance scores individually for each predictor, the function does not discard an entire row when values in the row are partially missing. For each variable, fsrftest uses all values that are not missing.

Example: 'UseMissing',true

Data Types: logical

Observation weights, specified as the comma-separated pair consisting of 'Weights' and a vector of scalar values or the name of a variable in Tbl. The function weights the observations in each row of X or Tbl with the corresponding value in Weights. The size of Weights must equal the number of rows in X or Tbl.

If you specify the input data as a table Tbl, then Weights can be the name of a variable in Tbl that contains a numeric vector. In this case, you must specify Weights as a character vector or string scalar. For example, if the weight vector is the column W of Tbl (Tbl.W), then specify 'Weights','W'.

fsrftest normalizes the weights to add up to one.

Data Types: single | double | char | string

Output Arguments

collapse all

Indices of predictors in X or Tbl ordered by predictor importance, returned as a 1-by-r numeric vector, where r is the number of ranked predictors.

If fsrftest uses a subset of variables in Tbl as predictors, then the function indexes the predictors using only the subset. For example, suppose Tbl includes 10 columns and you specify the last five columns of Tbl as the predictor variables by using formula. If idx(3) is 5, then the third most important predictor is the 10th column in Tbl, which is the fifth predictor in the subset.

Predictor scores, returned as a 1-by-r numeric vector, where r is the number of ranked predictors.

A large score value indicates that the corresponding predictor is important.

  • If you use X to specify the predictors or use all the variables in Tbl as predictors, then the values in scores have the same order as the predictors in X or Tbl.

  • If you specify a subset of variables in Tbl as predictors, then the values in scores have the same order as the subset.

For example, suppose Tbl includes 10 columns and you specify the last five columns of Tbl as the predictor variables by using formula. Then, score(3) contains the score value of the 8th column in Tbl, which is the third predictor in the subset.

Tips

  • If you specify the response variable and predictor variables by using the input argument formula, then the variable names in the formula must be both variable names in Tbl (Tbl.Properties.VariableNames) and valid MATLAB identifiers.

    You can verify the variable names in Tbl by using the isvarname function. The following code returns logical 1 (true) for each variable that has a valid variable name.

    cellfun(@isvarname,Tbl.Properties.VariableNames)
    If the variable names in Tbl are not valid, then convert them by using the matlab.lang.makeValidName function.
    Tbl.Properties.VariableNames = matlab.lang.makeValidName(Tbl.Properties.VariableNames);

Algorithms

collapse all

Univariate Feature Ranking Using F-Tests

  • fsrftest examines the importance of each predictor individually using an F-test. Each F-test tests the hypothesis that the response values grouped by predictor variable values are drawn from populations with the same mean against the alternative hypothesis that the population means are not all the same. A small p-value of the test statistic indicates that the corresponding predictor is important.

  • The output scores is –log(p). Therefore, a large score value indicates that the corresponding predictor is important. If a p-value is smaller than eps(0), then the output is Inf.

  • fsrftest examines a continuous variable after binning, or discretizing, the variable. You can specify the number of bins using the 'NumBins' name-value pair argument.

References

[1] Rasmussen, C. E., R. M. Neal, G. E. Hinton, D. van Camp, M. Revow, Z. Ghahramani, R. Kustra, and R. Tibshirani. The DELVE Manual, 1996.

[2] University of Toronto, Computer Science Department. Delve Datasets.

[3] Nash, Warwick J., ed. The Population Biology of Abalone (Haliotis Species) in Tasmania. 1: Blacklip Abalone (H. Rubra) from the North Coast and the Islands of Bass Strait. Technical Report/Department of Sea Fisheries, Tasmania 48. Taroona: Marine Research Laboratories, 1994.

[4] Waugh, S. "Extending and Benchmarking Cascade-Correlation." PhD Thesis. Computer Science Department, University of Tasmania, 1995.

[5] Lichman, M. UCI Machine Learning Repository. Irvine, CA: University of California, School of Information and Computer Science, 2013. http://archive.ics.uci.edu/ml.

Introduced in R2020a