Matrix: Wrong number of input argument(s): At least 2 expected

If I change the 1 line code than error is different. That line is

        model.state = [initial_state; initial_covariance(:)];

Whole code is

//Kalman_Filter_IF.sci
// The interface file for the generic Kalman Filter block.
// This block is externally activated to take measurements data.
// Takes the continuous positions velocity or accelerations as input, adds some 
// errors (noise, scale, nonlinearity, misalignment) and gives
// it as the output.

// Abbrev.
// IF: Inertial Frame, BF: Body Frame, 

// Model input: 
// 2x2 Real - State vector with position and velocity BF w.r.t. IF 
// 1 Real - Covariance vector

// Model output:
// 2x2 Real - Single axis gyro measurement output w.r.t. BF
// 1x1 Real - Error flag
// 0: No error
// 1: Positive saturation error
// 2: Negative saturation error

// Model states:
// 1x1 Real - Random walk error

// Model parameters:
// 2x2 Real - State Vector measurement unit vector w.r.t. BF (Unit vector)
// 1x1 Real - Noise mean (bias) of the sensors
// 1x1 Real - Noise stddev of the sensors
// 1x1 Real - Scale error
// 2x1 Real - Misalignment error 
// 1x1 Real - Random walk error
function [x, y, typ] = Kalman_Filter_IF(job, arg1, arg2)
    funcprot(0)
    x = [];
    y = [];
    typ = [];
    getInputFromSim = 1;
         initial_state = [0.0; 0.0];                  // 2*1 matrix
         initial_covariance = [1.0, 0.0; 0.0, 1.0];  // 2*2 matrix
    
    funcprot(0) // Use funcprot(0) to avoid this message it happen when we see inbuilt funcitons by system

    select job
        // Initialize with mask parameters if any
    case "plot" then
        standard_draw(arg1);
    case "getinputs" then
        [x, y, typ] = standard_inputs(arg1);

    case "getoutputs" then
        [x, y, typ] = standard_outputs(arg1);

    case "getorigin" then
        [x, y] = standard_origin(arg1);

    case 'set'
        // Assignments here comes after assignments of the execution with the "define" flag 
        x = arg1;  // This line is essential, otherwise gives "Invalid index." error.
        graphics = arg1.graphics;
        exprs = graphics.exprs;
        model = arg1.model;
        
        if getInputFromSim == 1 then
            exec("macros/Kalman_Filter_Utils.sci", -1); // Not sure about this line in MATLAB, comment it if not necessary
            title = "Set Parameter";
            labels = [  'Initial state [x_dot(velocity), x(position)] which is position in ECEF frame and velocity';
                        'Covariance which gives the uncertainty at position and velocity'];
  // Changed 'Stiffness' to 'Covariance'
            // In future we can add {'Accelerations', 'Orientations which is obation altitude control system'}
            types = list('mat', [2,2], 'mat', [2,2]);  // Changed 'vec' to 'mat' for Covariance
            endLoop = %f;
            disp(36)
            // Create a dialog for getting user inputs
            while ~endLoop do
                // Initialize dialog for parameters and initial state
                [ok, State, Covariance, exprs] = scicos_getvalue(title, labels, types, exprs);
                errMsg = [];

                // Check return set conditions of the initial state as per meetings
                if ~ok then
                    // depends on what type of data we are having
                    if condition1 // first condition of position and velocity
                        errMsg = [];
                    end 

                    if condition2 // second condition of velocity
                        errMsg = [];
                    end

                    if isempty(errMsg) then
                        
                        [model.rpar, isError] = EncodeRpar_KF(model.rpar, State, Covariance);
                        graphics.exprs = exprs;
                        x.graphics = graphics;
                        x.model = model;
                        endLoop = %t;
                        // Check parameters
                        // You can add additional checks for input parameters here if needed

                        // Accept inputs and save them
                    else
                        messagebox(errMsg);
                        error(errMsg, 20000);
                        endLoop = %f;
                    end
                else
                    warning('Kalman_Filter_IF: Set: Cancel!');
                    endLoop = %t;
                end
            end
        else
            graphics.exprs = exprs;
            x.graphics = graphics;
            x.model = model;
        end

        // Concatenate the covariance matrix to the state vector
        // model.state = [State; Covariance(:)];  // Use the concatenation operator ":" for the matrix
        disp(84)
        // Define block properties
    case 'define'then
        // Create object
        model = scicos_model();

        // Provide name and type
        model.sim = {'Kalman_Filter_sim', 5}; // what is case 4 and 5? https://help.scilab.org/scicos_model.html

        // Define inputs and outputs
        // One input with a variable-size "double" element
        model.in = 1; // it's only z,[-ve, complier will try to find appropriate size, 0 and +ve]
        // it has at first position occupies which means at first position is 1, 1 indicates that only one input value, see wriiten notes to understand
        model.in2 = 1;   // the second dimension of I/P whcih means all the data enter has one dimension data for example just z axis data while if it was 2 than we have input data of x and y if there is 3 then x,y,z so on. each column represents specific data. Same as above comment [https://help.scilab.org/scicos_model.html]
        model.intyp = 1;

        // One output with a single "double" element
        model.out = [2; 2];
        model.out2 = [1;2];
        model.outtyp = 1;
        // Set default parameter
        model.rpar= [];  // Use the covariance matrix as a parameter
        ipar=[1;1;1;1;2;3;1;3;1;1];

        // Set initial state
        model.state = [initial_state,matrix(initial_covariance)];   // Use the matrix concatenation operator ":" to convert matrix to column vector          model.state=[State;Covariance]
        model.state = [initial_state; initial_covariance(:)];
        model.rpar = initial_covariance;  // Use the covariance matrix as a parameter
        // Define block properties
        model.blocktype = 'c';
        model.dep_ut = [%t,%f]; // what does %t and %f indicates? true and false

        // Set block properties
        exprs = ['[0.0; 0.0]', '[1.0]']; // How do we get this value?
        x = standard_define([4 4], model, exprs);
        x.graphics.style = ['blockWithLabel;displayedLabel=Kalman_Filter'];
        // other cases other than set or define sections
    end
endfunction