# Using a set of data to determine values on the same line

1. Mar 16, 2014

### ka0ttic

Hey everyone,

Firstly, I apologize if there is a more suitable forum. I think this would be considered a calculus problem, but I'm not sure. I'm trying to develop a RF loss calculator for work that can calculate RF loss over a given length of cable or through a passive device at a given frequency. My math is not that great, so I'm hoping you guys can put me on the right track.

As an example, here's a manufacturer data sheet for a 2way coax splitter (rated 5-1000MHz):

Code (Text):

Frequency (MHz)                             Insertion Loss (dB/mV)
5                                                   4.2
40                                                  3.8
50                                                  3.8
450                                                4.3
550                                                4.4
750                                                4.6
870                                                4.9
1000                                              5.5

I would like to figure out the math that would allow me to find out the insertion loss based on any given frequency using that data to find it. For example, if I wanted to find out what the insertion loss would be at 300MHz, the data sheet goes from 50-450MHz with nothing in-between. Obviously, with the small amount of change, I could guess and be close enough, but if I'm going to write an application to do the math, I want it to be as accurate as possible.

I know that slope is involved but I'm not really sure about calculating slope for a curved line. I also am unsure of what I need to do once I figure out the slope to determine the insertion loss for a given freq.

Is anyone able to point me in the right direction? I'm totally for learning this myself, but I've had a hard time trying to do that, just because I don't necessarily know all the correct math lingo to research it.

Thanks for any help,
Aaron

2. Mar 16, 2014

### Vaedoris

What you need is regression analysis and someone with good knowledge of Statistics and experiment design.

3. Mar 16, 2014

### Stephen Tashi

Are you willing to use a different formula for each different type of data table?

If you plot the data in a table and it looks approximately like a line then you can use "linear interpolation".

If you encounter tables where the plot is very different than a line then there are other methods. At a glance, the example you gave looks non-linear. If you post some plots of type of data you encounter then perhaps someone will suggest good methods to use.