fscmrmr output only zero scores, is it possible?

fscmrmr function outputs only zero score for all features. I don't know if this is the expected behaviour.

[idx, scores] = fscmrmr(features,Y);
                results = [idx;scores];
                disp(results)
%features is a matrix. In each row there is one observation. In each column there is one feature. 
% Y is a logical vector (labels 1 and 0). It shows the same behaviour even if i use tables.                

[苑瑶 谢]

Have you solved this problem？I have the same questions in my dataset.

[Giovanni Pasini]

I didn't manage to solve the problem!

[bradley hovan]

Same issue, no luck thus far??

[John D'Errico]

Without seeing the specific data posed, it is impossible to know. Can it return zero scores, yes, I would bet that is the case. In fact, I am even positive that is the case, since in a random test case I just tried, that was true. But that just means you are probably misunderstanding how the code should be used.

help fscmrmr

 FSCMRMR Importance of features (predictors) for classification using MRMR algorithm.
    IDX=FSCMRMR(TBL,Y) ranks predictors for table TBL and class label Y.
    The ranks are learned using the Minimum Redundancy Maximum Relevance
    algorithm.
 
    TBL contains the predictor variables. Y can be any of the following:
       1. An array of class labels. Y can be a categorical array, logical
          vector, numeric vector, string array or cell array of character 
          vectors.
       2. The name of a variable in TBL. This variable is used as the
          response Y, and the remaining variables in TBL are used as
          predictors.
       3. A formula character vector such as 'y ~ x1 + x2 + x3' specifying
          that the variable y is to be used as the response, and the other
          variables in the formula are predictors. Any table variables not
          listed in the formula are not used.
 
    IDX=FSCMRMR(X,Y) is an alternative syntax that accepts X as an
    N-by-P matrix of predictors with one row per observation and one column
    per predictor. Y is the response and is an array of N class labels. 
 
    IDX is a 1-by-P vector for P predictors. IDX are indices of
    columns in X ordered by importance, meaning IDX(1) is the index of
    the most important predictor.  
 
    [IDX,SCORES]=FSCMRMR(...) also returns predictor scores SCORES, a
    1-by-P array for P predictors. SCORES have the same order as predictors
    in the input data, meaning SCORES(1) is the score for the first
    predictor. Large score indicates important predictor.
 
    [...]=FSCMRMR(X,Y,'PARAM1',val1,'PARAM2',val2,...) specifies optional
    parameter name/value pairs:
        'CategoricalPredictors' - List of categorical predictors. Pass
                         'CategoricalPredictors' as one of:
                           * A numeric vector with indices between 1 and P,
                             where P is the number of columns of X or
                             variables in TBL.
                           * A logical vector of length P, where a true
                             entry means that the corresponding column of X
                             or TBL is a categorical variable. 
                           * 'all', meaning all predictors are categorical.
                           * A string array or cell array of character
                             vectors, where each element in the array is
                             the name of a predictor variable. The names
                             must match variable names in table TBL.
                         Default: for a matrix input X, no categorical
                         predictors; for a table TBL, predictors are
                         treated as categorical if they are strings, cell
                         arrays of character vectors, logical, or unordered
                         of type 'categorical'.
        'ClassNames'   - Array of class names. Use the data type that
                         exists in Y. You can use this argument to select a
                         subset of classes out of all classes in Y.
                         Default: All class names in Y.
        'Prior'        - Prior probabilities for each class. Specify as one of:
                          * A character vector:
                            - 'empirical' determines class probabilities
                              from class frequencies in Y
                            - 'uniform' sets all class probabilities equal
                          * A vector (one scalar value for each class)
                          * A structure S with two fields: S.ClassProbs
                            containing a vector of class probabilities, and
                            S.ClassNames classes containing the class names and
                            defining the ordering of classes used for the
                            elements of this vector.
                         If you pass numeric values, FSCMRMR normalizes
                         them to add up to one. Default: 'empirical'
        'UseMissing'   - Logical value specifying whether missing values 
                         in the predictors must be used or discarded. When
                         'UseMissing' is set to true, FSCMRMR groups
                         together missing values and uses them to compute
                         mutual information between pairs of features and
                         between each feature and the response variable.
                         When 'UseMissing' is set to false, FSCMRMR omits
                         missing values in each feature while computing the
                         mutual information. Default: false
        'Weights'      - Vector of observation weights, one weight per
                         observation. FSCMRMR normalizes the weights to
                         add up to the value of the prior probability in
                         the respective class. Default: ones(size(X,1),1).
                         For an input table TBL, the 'Weights' value can be
                         the name of a variable in TBL.
        'Verbose'      - Verbosity flag, a nonnegative integer:
                          * 0  - FSCMRMR does not display any diagnostic
                                 messages (default).
                          * 1  - FSCMRMR displays major diagnostic info.
                          * >1 - FSCMRMR displays a lot of diagnostic info.
 
    Example:
        % Identify important predictors in the ionosphere dataset:
        load ionosphere
        [idx,scores] = fscmrmr(X,Y);
        bar(scores(idx))
        xlabel('Predictor rank')
        ylabel('Predictor importance score')

    Documentation for fscmrmr
       doc fscmrmr

A quick run of the sample test problem they supply in the help shows this:

load ionosphere
[idx,scores] = fscmrmr(X,Y)

idx = 1×34

5 4 1 7 24 29 12 3 34 33 15 21 9 32 31 27 25 8 13 19 30 6 23 10 11 17 16 14 2 18

scores = 1×34

0.0662 0 0.0214 0.0756 0.2905 0.0068 0.0345 0.0085 0.0128 0.0063 0.0023 0.0274 0.0079 0.0007 0.0145 0.0014 0.0018 0 0.0073 0 0.0141 0 0.0066 0.0310 0.0086 0 0.0098 0 0.0280 0.0072

Anyway, in order to understand what you did wrong to get the results you got, you need to show the data you tried to pass into that code.