Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                
Scribd Logo
Upload

Welcome to Scribd!

  • 73 views

    Error Propagation

    Uploaded by

    Tanzid Sultan
    The document discusses error propagation and how to calculate the error in a variable that depends on other variables with errors by taking the partial derivative of the variable with respect to each variable it depends on and multiplying by the error of that variable. It also discusses estimating the error in the slope of a best fit line by drawing lines above and below the best fit line that still fit within the error bars and calculating the error as the difference between the maximum and minimum slopes divided by two.

    Copyright:

    © All Rights Reserved
    Download as PPTX, PDF, TXT or read online from Scribd

    Error Propagation

    Uploaded by

    Tanzid Sultan
    73 views25 pages
    The document discusses error propagation and how to calculate the error in a variable that depends on other variables with errors by taking the partial derivative of the variable with respect to each variable it depends on and multiplying by the error of that variable. It also discusses estimating the error in the slope of a best fit line by drawing lines above and below the best fit line that still fit within the error bars and calculating the error as the difference between the maximum and minimum slopes divided by two.

    Copyright

    © © All Rights Reserved

    Available Formats

    PPTX, PDF, TXT or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    The document discusses error propagation and how to calculate the error in a variable that depends on other variables with errors by taking the partial derivative of the variable with respect to each variable it depends on and multiplying by the error of that variable. It also discusses estimating the error in the slope of a best fit line by drawing lines above and below the best fit line that still fit within the error bars and calculating the error as the difference between the maximum and minimum slopes divided by two.

    Copyright:

    © All Rights Reserved
    Download as PPTX, PDF, TXT or read online from Scribd
    Download as pptx, pdf, or txt
    73 views25 pages

    Error Propagation

    Uploaded by

    Tanzid Sultan
    The document discusses error propagation and how to calculate the error in a variable that depends on other variables with errors by taking the partial derivative of the variable with respect to each variable it depends on and multiplying by the error of that variable. It also discusses estimating the error in the slope of a best fit line by drawing lines above and below the best fit line that still fit within the error bars and calculating the error as the difference between the maximum and minimum slopes divided by two.

    Copyright:

    © All Rights Reserved
    Download as PPTX, PDF, TXT or read online from Scribd
    Download as pptx, pdf, or txt
    You are on page 1of 25

    Propagation of Error

    . . . and all about error


    What is error (a.k.a. “uncertainties”)?
    Error is a deviation in measured value from true value.

    True value of the measurement

    Error in that measurement


    What is error (a.k.a. “uncertainties”)?

    We notate error by

    For example, if we measured a distance, y, of 5mm with an error of 0.5mm, we


    would write
    Common Types of Errors
    ● Precision Error: Error due to limited precision of the measuring instrument.
    As a rule of thumb, a precision error is taken to be half the smallest unit that
    the instrument is capable of measuring. E.g. A meter stick can’t measure
    anything less than . So the error in distances measured using a meter stick
    can be taken to be .

    ● Human Error: Error due to difference between “perceived” measured value


    and true value. E.g. a stopwatch time measurement will have error due to
    human reaction delays. The size of this error depends on the circumstances
    of the measurement. Need to make an educated guess when estimating the
    numerical value this error.
    Propagation of Error
    Imagine we have some variable f which depends on x. If x has some error, then
    how will that error affect f?
    f

    x
    Propagation of Error
    Imagine we have some variable f which depends on x. If x has some error, then
    how will that error affect f?
    f

    x
    x1
    Propagation of Error
    Imagine we have some variable f which depends on x. If x has some error, then
    how will that error affect f?
    f

    f(x1)

    x
    x1
    Propagation of Error
    Imagine we have some variable f which depends on x. If x has some error, then
    how will that error affect f?
    f
    This new error is

    𝛿𝑓
    f(x1)

    x
    x1
    Propagation of Error
    How can we calculate the error in f due to x?
    Propagation of Error
    How can we calculate the error in f due to x?
    𝛿𝑓

    𝛿𝑥
    Propagation of Error

    How can we calculate the error in f due to x?


    𝛿𝑓

    For example: 𝛿𝑥
    Propagation of Error

    How can we calculate the error in f due to x?


    𝛿𝑓

    For example: 𝛿𝑥

    and
    Propagation of Error
    What if we have >1 variables? Ie. f(x,y,.......,z)

    What are these?


    Propagation of Error
    What if we have >1 variables? Ie. f(x,y,.......,z)

    What are these?

    Partial Derivatives work the same as regular derivatives except


    you pretend every variable is a constant except for “x”.

    So they are even easier!


    Estimating error in slope of a best fit line

    Let be the slope of the best fit


    line.
    Estimating error in slope of a best fit line

    Let be the slope of the best fit


    line. To estimate the error in ,
    draw two other lines that over-
    and under-estimate the slope,
    while still fitting within all the
    error bars.
    Estimating error in slope of a best fit line

    Let be the slope of the best fit


    line. To estimate the error in ,
    draw two other lines that over-
    and under-estimate the slope,
    while still fitting within all the
    error bars.
    Estimating error in slope of a best fit line

    Let be the slope of the best fit


    line. To estimate the error in ,
    draw two other lines that over-
    and under-estimate the slope,
    while still fitting within all the
    error bars.
    Estimating error in slope of a best fit line

    Then, estimate the slope


    error as

    𝑎 𝑚𝑎𝑥 − 𝑎𝑚𝑖𝑛
    𝛿 𝑎=
    2

    𝛿𝑎

    𝑎 𝑚𝑖𝑛 𝑎 𝑎 𝑚𝑎𝑥

    where and are the slopes of the two lines that over- and under-estimate the slope of the
    best fit.
    What I expect
    1. For your quantity of interest, you will propagate the error in the variable which
    contributes most to the error.

    Super reasonable question: “How do we know which variable contributes most to


    the error without actually calculating it?”
    What I expect
    1. For your quantity of interest, you will propagate the error in the variable
    which contributes most to the error.

    Super reasonable question: “How do we know which variable contributes most to


    the error without actually calculating it?”

    Answer: You don’t. You will have to reasonably guess and justify your choice!
    What I expect
    1. For your quantity of interest, you will propagate the error in the variable
    which contributes most to the error.

    Super reasonable question: “How do we know which variable contributes most to


    the error without actually calculating it?”

    Answer: You don’t. You will have to reasonably guess and justify your choice!

    Example: If you have a quantity that depends on three variables which have
    errors , and you’ve determined that variable contributes most to the error, then
    What I expect
    1. For your quantity of interest, you will propagate the error in the variable which
    contributes most to the error.
    2. For all values that you directly measure, you should also record the error
    associated with the measurement
    What I expect

    1. For your quantity of interest, you will propagate the error in the variable which
    contributes most to the error.
    2. For all values that you directly measure, you should also record the error
    associated with the measurement.

    Example: Lets say that you’ve performed an experiment in which you measured
    distance () using a meterstick and time () using a stopwatch. Then you need to
    state the values of the uncertainties and .
    What I expect

    1. For your quantity of interest, you will propagate the error in the variable which
    contributes most to the error.
    2. For all values that you directly measure, you should also record the error
    associated with the measurement.
    3. Include error bars in all your plots. Error bars encapsulate the error in each
    data point.

    You might also like