Domain Error when finding an exponential regression

[Mark Arguijo]

Hi Everybody!

Just a quick question....  I'm trying to find an exponential regression from the following data

x     y

-2    -1/2

-1     0

0      1

1      3

When I enter the data in a Lists & Spreadsheet and try to do a regression, I get "Domain Error.  An argument must be in a specified domain."

What am I doing wrong?

Mark

[CHa...@cattlv.wnyric.org]

1st, make sure that you used the negative not subtraction
2nd, make sure after you typed the last data element you actually hit "enter"



Cathy Haynes
7th grade accelerated math
Algebra 2
Pre-Calculus
Cattaraugus Little Valley Central School
25 N Franklin Street
Cattaraugus, NY  14719
phone:  (716) 257-3483 ext. 1135



In mathematics you don't understand things. You just get used to them.
Johann von Neumann
US (Hungarian-born) computer scientist, mathematician (1903 - 1957)

-----tins...@googlegroups.com wrote: -----

To: tinspire <tins...@googlegroups.com>
From: Mark Arguijo
Sent by: tins...@googlegroups.com
Date: 04/18/2016 02:15PM
Subject: [tinspire] Domain Error when finding an exponential regression

--
--
To post to this group, send email to tins...@googlegroups.com
To unsubscribe send email to tinspire+u...@googlegroups.com
For more options, visit this group at
http://groups.google.com.au/group/tinspire?hl=en-GB?hl=en-GB
The tns documents shared by group members are archived at
https://sites.google.com/site/tinspiregroup/classroom-news/welcome-abouttime

---
You received this message because you are subscribed to the Google Groups "tinspire" group.
To unsubscribe from this group and stop receiving emails from it, send an email to tinspire+u...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

---

Spam
Not spam
Forget previous vote

[Travis Bower]

Try (0.001, 1) just to see if you can get some traction.

I recall sometimes getting that error when I included the y-intercept....or that is what I thought was the issue.

OS4.2 right?

--

--
To post to this group, send email to tins...@googlegroups.com
To unsubscribe send email to tinspire+u...@googlegroups.com
For more options, visit this group at
http://groups.google.com.au/group/tinspire?hl=en-GB?hl=en-GB
The tns documents shared by group members are archived at
https://sites.google.com/site/tinspiregroup/classroom-news/welcome-abouttime

---
You received this message because you are subscribed to the Google Groups "tinspire" group.
To unsubscribe from this group and stop receiving emails from it, send an email to tinspire+u...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

--

	Travis Bower Math Teacher Dos Pueblos High School tbower@sbunified.org www.dphs.org www.sbunified.org 805.968.2541 x4538

---

CONFIDENTIALITY NOTICE: This is a transmission from the Santa Barbara Unified School District and may contain privileged and confidential information. It, and any attachments, are intended only for the addressee(s) . If you are not the intended recipient, you are hereby notified that any review, dissemination, distribution or duplication of this communication and its attachments is strictly prohibited by applicable state and federal law. If you are not the intended recipient, please contact the sender by reply email and destroy all copies of the original message and attachments.

Santa Barbara Unified School District - "Every child, every chance, every day."

[Mark Arguijo]

I did use the negative key and enter was pressed after the last element. 

I still get the error. 

Mark Arguijo

T^3 Regional Instructor

Email:  mrar...@gmail.com

Mobile:  956-245-6090

Sent from my iPhone

[Dennis Donovan]

I think that error message is misleading at best, it really should say that the data need to be in a specific range.  Namely y>0.

Looking at the Reference Guide, the result of exponential regression is a function of the form a*b^x, this means no vertical translation, therefore all y coordinates must be > 0.

If you translate your data up 1 unit, the regression will work.

My guess is that the calculator is not truly doing an exponential regression, rather it is doing a transformation of the data by taking the natural logarithm of the y coordinates, then doing a linear regression and finally transforming that linear regression back to an exponential function of the form a*b^x.  This is why negative y-values cause a problem.

Dennis

---

From: Mark Arguijo <mrar...@gmail.com>
To: tinspire <tins...@googlegroups.com>
Sent: Monday, April 18, 2016 2:17 PM

Subject: [tinspire] Domain Error when finding an exponential regression

[Mark Arguijo]

Did that, Travis. 

Still no luck and yes, this is on OS 4.2. 

Mark Arguijo

T^3 Regional Instructor

Email:  mrar...@gmail.com

Mobile:  956-245-6090

Sent from my iPhone

[Mark Arguijo]

When I remove the data with negative y values, I still get the same errors. 

Mark Arguijo

T^3 Regional Instructor

Email:  mrar...@gmail.com

Mobile:  956-245-6090

Sent from my iPhone

[Travis Bower]

I like DD's thoughts about the translation.  Did that work?

[Landy Godbold]

Note that DD said y > 0. Did you also remove the y = 0 data?

----- Original Message -----
From: "Travis Bower" <tbo...@sbunified.org>
To: "tinspire" <tins...@googlegroups.com>

> *Travis Bower*

> Math Teacher
> Dos Pueblos High School

> tbo...@sbunified.org
> www.dphs.org
> www.sbunified.org
> 805.968.2541 x4538
> <https://www.facebook.com/sbunified> <https://twitter.com/sbunified>
> <https://www.youtube.com/channel/UCtBzORVgEnNrp6IJyTP5_bA>
> <http://www.pinterest.com/sbunified/>
> <https://itunes.apple.com/us/app/santa-barbara-usd/id915427103?mt=8&uo=4>
>
>
> ------------------------------

*Travis Bower*

Math Teacher
Dos Pueblos High School

tbo...@sbunified.org
www.dphs.org
www.sbunified.org
805.968.2541 x4538
<https://www.facebook.com/sbunified> <https://twitter.com/sbunified>
<https://www.youtube.com/channel/UCtBzORVgEnNrp6IJyTP5_bA>
<http://www.pinterest.com/sbunified/>
<https://itunes.apple.com/us/app/santa-barbara-usd/id915427103?mt=8&uo=4>

--

------------------------------

[Mark Arguijo]

Well, that did give me a regression but the wrong one since there were only 2 data points left.

I have to mention that this is a review question for an upcoming state exam and those 4 pieces of data are all that's given.

Any other ideas?

Mark Arguijo


Sent from my iPhone

[Travis Bower]

Travis Bower

Math Teacher Dos Pueblos High School

tbower@sbunified.org www.dphs.org www.sbunified.org 805.968.2541 x4538

[Mark Arguijo]

Thanks, Dennis. 

I translated the y coordinates one unit up and was able to finally get a regression, f(x)=2*2^x, and then subtract one to get it to fit the data. 

But there has to be something easier, shouldn't there be?

Mark Arguijo

Sent from my iPhone

On Apr 18, 2016, at 1:40 PM, 'Dennis Donovan' via tinspire <tins...@googlegroups.com> wrote:

[Mark Arguijo]

I ended up doing that (see my reply to Dennis). That's the best workaround I can see at the moment. 

Mark Arguijo

Sent from my iPhone

On Apr 18, 2016, at 2:23 PM, Travis Bower <tbo...@sbunified.org> wrote:

<image.png>

<expmark.tns>

[Travis Bower]

But there has to be something easier, shouldn't there be?

Maybe yes and maybe no.....

Maybe the test makers want to thwart blindly typing data into a device?

Maybe Ti should consider y=a*b^(x-h) +k?  But in the mean time, they should specify on the device what regression it is finding...not just ellipses.  The Catalog did not help.  Maybe it is somewhere else?

	Travis Bower Math Teacher Dos Pueblos High School tbower@sbunified.org www.dphs.org www.sbunified.org 805.968.2541 x4538

[Landy Godbold]

That's sad. It means that the person/committee writing the questions do not know the math they are testing! So your students will first have to guess at what mathematics was _intended_ then apply what they know. Best of luck.

[Dennis Donovan]

The problem is knowing the right translation to make the regression work, here it had to be 1 because the horizontal asymptote for the original data is y=-1, but how is a student going to know that.

I'd like to see the original problem, what is it asking for, is it multiple choice or free response?

---

From: Mark Arguijo <mrar...@gmail.com>

To: tins...@googlegroups.com
Sent: Monday, April 18, 2016 3:25 PM
Subject: Re: [tinspire] Domain Error when finding an exponential regression

[Travis Bower]

also, please try Vernier's data analysis...gotta teach right now...

Travis Bower

Math Teacher Dos Pueblos High School

tbower@sbunified.org www.dphs.org www.sbunified.org 805.968.2541 x4538

[Travis Bower]

oops, I could not remember if translation was an option....VDQ gives some other options.

Also, here is where we see some equations.

[Sean Bird]

Yes, the TI-Nspire's Vernier DataQuest application does make it easy to enter the x & y values in its table that is set up for x & y. Then press the graph icon at the bottom or press Menu > View > Graph. (If you don't like the values connected, use Menu > Options > Points Options)

Just under the Curve Fit in the Menu > Analyze is Model. You can write your own model or start from one provided. We know if we want to model this data that it isn't linear or quadratic. Does this and the image above provide enough information to explore?

Yes, some regressions will give domain errors because the values cannot be modeled without translation.

Enjoy,

Sean Bird

[Mark Arguijo]

Thanks for the tips, Sean. 

I had tried the data quest app but still couldn't get a regression. The Model option does allow for students to explore and match a model to their data. 

Thanks again!

Mark Arguijo

Sent from my iPhone

On Apr 19, 2016, at 7:23 AM, Sean Bird <covena...@gmail.com> wrote:

Yes, the TI-Nspire's Vernier DataQuest application does make it easy to enter the x & y values in its table that is set up for x & y. Then press the graph icon at the bottom or press Menu > View > Graph. (If you don't like the values connected, use Menu > Options > Points Options)

<image.png>
<image.png>

Enjoy,

--