How can I perform exponential regression analysis on an uploaded data set?

Questions
1

I have two columns, ten rows. Column A contains x values; column B contains y values. These are in a CSV file, which uploads fine. I want to use “exponential fit” against this data.

To be clear, I can upload the CSV file. Conversely I can manually input the data and use “exponential fit” in the command line with that data. I want to combine these two steps: I want to use “exponential fit” against a large CSV.

Thanks for any suggestions.

Bruce

2

may be you can look at NonlinearModelFit ? Something like

model = NonlinearModelFit[myData, a*Exp[b*x], {a,b}, x];
model["BestFitParameters"]

see https://reference.wolfram.com/language/ref/NonlinearModelFit.html for more examples

3

Thank you. Your suggestion is probably good, but I’m looking for commands I can enter into the Wolfram command line, or some other way to run “exponential fit” against my uploaded data.

4

When you say “Wolfram Command Line”, do you mean “Wolfram Alpha”?

5

Yes. Sorry for the ambiguity. I mean the search field (with the equal-sign submit button) at wolframalpha.com.

6

Here is a list of examples of what kinds of things Wolfram|Alpha does with uploaded data:

http://www.wolframalpha.com/examples/DataInput.html

In particular it looks like this is an example of a nonlinear regression:

Wolfram|Alpha is really kinda loose and does exploratory stuff for you. If you have something specific in mind, I would use Mathematica Online and the code provided above.

7

Thank you. That helps, although I can’t see what my options are for the words following the uploaded data. (My original question was about syntax and available commands/expressions/functions.) I’ll look for “time series” and see what related options I might have.

I’m a high school teacher, and I don’t want to spend too much time teaching a programming language to the students. I want to take the data and process it, with a little more power than Desmos, a little more availability than the commands on a TI graphing calculator. I hope that makes sense.

8

Obviously, I haven’t made myself clear. I’ll try again.

I can do this: exponential fit {(0,185),(1,170),(2,160),(3,152),(4,145),(5,140),(6,135),(7,130),(8,126)} - Wolfram|Alpha

At www.wolframalpha.com, I can click a File Upload icon and upload a CSV with two columns (containing the data shown in the link above). The search field now contains an icon representing my uploaded file. I want to type “exponential fit” after this file icon and get the results shown in the link above.

If I do type “exponential fit” and click the Equal Sign icon, I get this error message: Using closest Wolfram|Alpha interpretation: exponential

Thanks for your help.

Bruce

9

Here are some examples where you enter the data directly into the bar:

exponential fit of {1,2,4,7,17}

exponential fit of {{0,1},{2,2},{4,4},{6,7},{8,17}}

exponential fit of {1,2,4,7,17} - Wolfram|Alpha

There are no examples showing how this can be done with a dataset. I am not aware of any. It’s very possible that the Wolfram|Alpha doesn’t recognize that query yet for datasets. If a query doesn’t work that you expect to, you can contact the Wolfram|Alpha team and leave them a suggestion at the bottom of the results page. This helps them know where to focus the future developement of Wolfram|Alpha.