solving transition probabilities

Dear All,

I have written a program for computing transition probabilities.

the transitional probability from s1 ->s2

1 compute total no number of successive events going from s1 -> s2 divided by the total number of s1 events. which is describe in following equation

Transitional probability s1 -> s2 = Pr[Xt+1 = s2|Xt = s1].

But I have error in calculating . Couldn't find where is error in my code

error is here :

Error in ==> statv2 at 4
if (x(1) <= 5)

??? Output argument "state_current" (and maybe others) not assigned during call to "C:\Markov
Chain Programs\Matlab CPU program - Copy (2)\statv2.m>statv2".

Error in ==> modelMarkov_v2 at 20
       [previousState currentState] = statv2([data(i-1,1) data(i,1)]);

Here is code for calculation transition probabilities

      reading = [77;66.9016000000000;67.0833000000000;66.1000000000000;66.3115000000000;66.3934000000000;86.8525000000000;78.7288000000000;66.4918000000000;66.7833000000000;67.3667000000000;66.7541000000000;68.2034000000000;67.5833000000000;65.7869000000000;66.4068000000000;85.3443000000000;90.9667000000000;66.3667000000000;67.9500000000000;69.7833000000000;67.6441000000000;66.9667000000000;66.1167000000000;65.6721000000000;67.4576000000000;69.1167000000000;96.6333000000000;66.4918000000000;66.3443000000000;66.9830000000000;78.5410000000000;67.3729000000000;66.5574000000000;66.5667000000000;66.8644000000000;68.3934000000000;110.610000000000;66.8500000000000;66.9167000000000;70.9508000000000;68.1695000000000;67.2667000000000;];

        data = reading ;
        transFreq = zeros(22,22);
        transCount = zeros(22,22);

        % calculating transion count in data

        i = 1;
        for i = 1:length(data)

           if (i < 2)
               [previousState currentState] = statv2([data(i,1) data(i,1)]);
               transCount(previousState, currentState) = transCount(previousState,currentState) + 1;
           end
           if (i >= 2)
               data(i-1,1);
               data(i,1);
               [previousState currentState] = statv2([data(i-1,1) data(i,1)]);
               transCount(previousState, currentState) = transCount(previousState, currentState) + 1;
            end
        end

        % Calculate Transistion Frequencies
        for j = 1:22
        transFreq(j,:) = transCount(j,:)/ sum(transCount(j,:));
        end

I have a function : statv2 to calculate previous state and current state

        function [state_prev,state_current] = statv2(x)

        % for previous state
        if (x(1) <= 5)
            state_prev = 1;
        end
        if (x(1) > 5 &&  x(1) <= 10)
            state_prev = 2;
        end
        if (x(1) > 10 &&  x(1) <= 15)
            state_prev = 3;
        end
        if (x(1) > 15 &&  x(1) <= 20)
            state_prev = 4;
        end
        if (x(1) > 20 &&  x(1) <= 25)
            state_prev = 5;
        end
        if (x(1) > 25 &&  x(1) <= 30)
            state_prev = 6;
        end
        if (x(1) > 30 &&  x(1) <= 35)
            state_prev = 7;
        end

        if (x(1) > 35 &&  x(1) <= 40)
            state_prev = 8;
        end

        if (x(1) > 40 &&  x(1) <= 45)
            state_prev = 9;
        end

        if (x(1) > 45 &&  x(1) <= 50)
            state_prev = 10;
        end

        if (x(1) > 50 &&  x(1) <= 55)
            state_prev = 11;
        end

        if (x(1) > 55 &&  x(1) <= 60)
            state_prev = 12;
        end

        if (x(1) > 60 &&  x(1) <= 65)
            state_prev = 13;
        end

        if (x(1) > 65 &&  x(1) <= 70)
            state_prev = 14;
        end

        if (x(1) > 70 &&  x(1) <= 75)
            state_prev = 15;
        end
        if (x(1) > 75 &&  x(1) <= 80)
            state_prev = 16;
        end
        if (x(1) > 80 &&  x(1) <= 85)
            state_prev = 17;
        end
        if (x(1) > 85 &&  x(1) <= 90)
            state_prev = 18;
        end
        if (x(1) > 90 &&  x(1) <= 95)
            state_prev = 19;
        end
        if (x(1) > 95 &&  x(1) <= 100)
            state_prev = 20;
        end
        if (x(1) > 100 &&  x(1) <= 105)
            state_prev = 21;
        end
        if (x(1) > 105 &&  x(1) <= 110)
            state_prev = 22;
        end

        % for current state
        if (x(2) <= 5)
            state_current = 1;
        end
        if (x(2) > 5 &&  x(2) <= 10)
            state_current = 2;
        end
        if (x(2) >10 &&  x(2) <= 15)
            state_current = 3;
        end
        if (x(2) > 15 &&  x(2) <= 20)
            state_current = 4;
        end
        if (x(2) > 20 &&  x(2) <= 25)
            state_current = 5;
        end
        if (x(2) > 25 &&  x(2) <= 30)
            state_current = 6;
        end
        if (x(2) > 30 &&  x(2) <= 35)
            state_current = 7;
        end

        if (x(2) > 35 &&  x(2) <= 40)
            state_current = 8;
        end

        if (x(2) > 40 &&  x(2) <= 45)
            state_current = 9;
        end

        if (x(2) > 45 &&  x(2) <= 50)
            state_current = 10;
        end

        if (x(2) > 50 &&  x(2) <= 55)
            state_current = 11;
        end

        if (x(2) > 55 &&  x(2) <= 60)
            state_current = 12;
        end

        if (x(2) > 60 &&  x(2) <= 65)
            state_current = 13;
        end

        if (x(2) > 65 &&  x(2) <= 70)
            state_current = 14;
        end

        if (x(2) > 70 &&  x(2) <= 75)
            state_current = 15;
        end
        if (x(2) > 75 &&  x(2) <= 80)
            state_current = 16;
        end
        if (x(2) > 80 &&  x(2) <= 85)
            state_current = 17;
        end
        if (x(2) > 85 &&  x(2) <= 90)
            state_current = 18;
        end
        if (x(2) > 90 &&  x(2) <= 95)
            state_current = 19;
        end
        if (x(2) > 95 &&  x(2) <= 100)
            state_current = 20;
        end
        if (x(2) > 100 &&  x(2) <= 105)
            state_current = 21;
        end
        if (x(2) > 105 &&  x(2) <= 110)
            state_current = 22;
        end

Could please help me how can I resolve the error

Many Thanks in advance

Answer 1

[Walter Roberson]

You have a data point 110.610000000000 which is beyond the limits that you test for, 110, so no state gets assigned in that situation.

Have you considered the health benefits of using histc() ?