# Thread: Extrapolation formulae

1. ## Extrapolation formulae

Hi,

Can you help me out?

I've data as follows:
X Y
8 5.93
10 5.87
12 5.84

I want to extapolate the graph and find out the value where this graph cuts the Y-axis i.e. values for 0,2,4 & 6.

2. Originally Posted by kshrini
Hi,

Can you help me out?

I've data as follows:
X Y
8 5.93
10 5.87
12 5.84

I want to extapolate the graph and find out the value where this graph cuts the Y-axis i.e. values for 0,2,4 & 6.
Put the data into Excel (or some other spread sheet) and plot it as a scatter plot. Select the line and right click and select add trend line, select linear for the type of trend.

Now right click on the trend line and select format trend line, click on display equation, also setup forcast to forcast backwards by 8 units.

Now read all you need off of the plot.

The attachment shows what you should have

RonL

3. ## Extrapolation formulae

Hi,

Thanks for reply. But I want to program it in Viusal Basic. So I want the formula.

Thanks once again.
Shrinivas

4. Originally Posted by kshrini
Hi,

Thanks for reply. But I want to program it in Viusal Basic. So I want the formula.

Thanks once again.
Shrinivas
The Excel help system entry for trend line will give you the equations
for the trend line calculation.

RonL

5. ## Extrapolation formulae

Hi,

Thanks for reply. My data is changing, as I'm acquiring it from instrument. So extrapolation will be done at the time of data acquisition.

If you have the formula then I can manage it programatically.

Shrinivas

6. Originally Posted by kshrini
Hi,

Thanks for reply. My data is changing, as I'm acquiring it from instrument. So extrapolation will be done at the time of data acquisition.

If you have the formula then I can manage it programatically.

Shrinivas
See here for the equations.

Summarising these:

Put

$s_{x,y}=sum(x_i y_i)$

$s_{x}=sum(x_i)$

$s_{y}=sum(y_i)$

$s_{x,x}=sum(x_i^2)$

$s_{y,y}=sum(y_i^2)$

Then:

$m=\frac{n s_{x,y}-s_x s_y}{n s_{x,x}-s_x^2}$,

where $n$ is the number of data points.

and:

$b=\frac{s_y-m s_x}{n}$.

Then the regression line is $y=m x+b$ and $y=b$ when $x=0$.

RonL

7. ## Extrapolation formulae

Hi,

Thanks a lot. This is perfectly OK if data has linear relation. But if the data is exponential then how to extrapolate it?

Shrinivas