Explore BrainMass
Share

Explore BrainMass

    Matlab normal equations model to analyze response

    This content was COPIED from BrainMass.com - View the original, and get the already-completed solution here!

    (See attached file for full problem description)

    ---
    This model is used to analyze the body's response to a bolus injection of this antibiotic:
    B(t) = B0e(-k*t)
    Using this model, and data provided for the average plasma concentration data in ug/ml with time in hours :
    a) perform linear regression on this penicillin clearance using the normal equations
    b) plot the concentration data and re-plot the log of the concentration data to show that it appears linear.
    c) Determine initial concentration, B0, and elimination rate constant, k

    Concentration (ug/ml) Time (min)
    89 22
    60 44
    30 88
    15 132
    7 176

    a) for linear regression
    ln( B ) = ln( B0 )- Kt

    I'm not sure how to use the normal equations in MATLAB to plot this? We're not supposed to use polyfit function... I think you have to make a 2 column matrix with a column of 1's and a column of the time?
    ---

    (See attached file for full problem description)

    © BrainMass Inc. brainmass.com April 3, 2020, 2:53 pm ad1c9bdddf
    https://brainmass.com/computer-science/matlab/49820

    Attachments

    Solution Preview

    Answer:

    From what you write I understand that you are to perform linear regression for the logarithmic fit: ln(B) = ln(B_0) - Kt

    If you find out otherwise, please tell me, and I shall try to help, but now I proceed with the stated understanding.

    BTW, the ln(B) = ln(B_0) - Kt fit would do exactly the same as the polyfit (x, y, 1) function of MatLab, but here you are apparently requested to do "manually" what the polyfit does.

    First I shall try to explain about the normal equations and linear regressions.
    For convenience, let us rename x_1 = ln(B_0) and x_2 = K.
    Then the equations based on the input data which you are supposed to try to fit as best you can are:

    1 * x_1 - 22 * x_2 = ln(89)
    1 * x_1 - 44 * x_2 = ln(60)
    1 * x_1 - 88 * x_2 = ln(30)
    1 * x_1 - 132 * x_2 = ln(15)
    1 * x_1 - 176 * x_2 = ln(7)

    You see here 5 equations for only 2 unknowns. Obviously it is not likely possible to find a solution that will make all these equations true. Instead, what ...

    $2.19

    ADVERTISEMENT