I need to use some technology to solve a problem, but unfortunately I have no idea which program I could use.
Basically, I just have some data points and need a program which can generate some functions that fit these data points as closely as possible. Possibly (probably) I could also use my TI-84 calculator, but I would prefer a computer program. Can maybe someone recommend me a program that is easy to use (and preferably free/cheap) and that can generate functions if I enter my data? That's all it needs to be able to do!
Thank you very much for your help, and sorry if this question is too easy, but I never really used technology before (except of a calculator) and have no idea which program would be good for that. (Hopefully you can recommend me something other than Excel)
February 22nd 2008, 01:48 PM
What's wrong with your calculator or Excel?. They work just fine.
On a TI, just enter the data and execute the regression you need.
If you're no good with tech, then now is the time to learn. We all had to at some point.
Post some data and I will be glad to show you how to enter it into Excel to find a regression equation, graph, or whatever.
February 23rd 2008, 11:59 AM
The problem with excel is that I do not have excel but only Open Office Calc, which, however, is very similar to excel. I have 10 x-values and 10 corresponding y-values and now I want to create a function that represents this relationship (ideally, I want to have a couple of different functions and then I want to determine which of them is the best). If you could tell me how I can do that with excel (I will then try to do the equivalent steps with my program) or with the TI I would appreciate it very much. Just tell me where I need to click and I'll enter my data. Thanks a lot!
February 23rd 2008, 12:15 PM
Post your 10 x and 10 y values. It would be easier if I could give you a screen capture off my TI-92. If you run a linear regression, polynomial regression, etc. the closer your R^2 value is to 1, the more accurate the regression analysis.
February 23rd 2008, 12:24 PM
Enter x values in column A
Enter y values in column B
Click on 'Insert'
'XY Scatter' plot
highlight your data and your graph should appear.
Click on 'Chart' in the toolbar
Then click on 'Add Trendline'
You will see 'polynomial'(among others). You can set it to 2nd degree to 6th degree.
Click on 'Options' and check the boxes that read "Display equation on graph" and "display R^2 value on graph".
That ought to be about it. Good luck. Remember, the closer your R^2 is to 1, the better.
March 3rd 2008, 12:46 PM
Thanks a lot for your help. Unfortunately it didn't work out with open office, so I downloaded excel 2003. There, however, it also doesn't work. I managed to graph the points, but there is no "add trendline". Can you please help me?
This is what I got, where do I need to click now? Thanks a lot!
March 3rd 2008, 01:29 PM
I can not read the data from the graph. It's too blurry. Let me know your points and I can find your equation. It would also be easier to show you how to proceed. I have asked before and you do not appear to want to give them to me. Does it have something to do with national security?.
March 3rd 2008, 01:47 PM
No, it doesn't, I just want to do the calculations myself. It's a matter of academic honor, I don't want you to do all the work for me, I want to find the equation myself. But okay, that's the data:
Please let me know how it exactly it works and where I need to click in excel 2003. I looked for "add trendline" in google, but all explanations I could find didn't work with my version of excel. Thanks very much for your help :)
March 3rd 2008, 02:00 PM
I was just teasing about the national security thing. (Giggle)
If you click on your graph and then click on 'chart' in the toolbar you should see 'add trendline'. If it's not there I don't know what to say. If you're using Vista, who knows what may have happened to it.
But, using your data, I arrived at an equation of
As you can see, R^2 is very close to 1. That is a good best-fit line.
Excel goes up to a polynomial of degree 6. The accuracy for a degree 6 wasn't that much greater.
March 3rd 2008, 02:10 PM
I cannot find this trendline thingy. Damn. Which excel version do you have? Do you know if there is a way to download a trial version of it? And also, could you explain to me what exactly this R^2 means and how it is calculated? Anyway, thank you very much for your help!
March 3rd 2008, 03:43 PM
I don't know what may be the matter. Try the 'help' option on Excel itself.
You know, calculators do this too. You don't have to use Excel.
Below is a screen capture from my TI-92 using your data.
R^2 is the ratio of the explained variation to the total variation.
In other words, in your case, the R^2 is .9921. That means that 99.21% of the variation in y can be explained by x. The other wee bit is just due to sampling error or some other variables.
March 3rd 2008, 08:12 PM
Thanks, I guess I'll do it on the calculator. Can you explain me which buttons I need to push?