Linreg through origin?

[Al Coons]

Am I correct that there is no way to force a linear regression through the origin on the Nspire?

Thanks,

Al

[Sean Bird]

Input the data in the Vernier DataQuest app and Analyze the data there. You can do Curve Fit, A: Natural Exponential a*e^(-c*x) or B:Proportional, y=a*x

These are two bonus regressions that you can't do in any other Nspire app.

- Sean Bird

[Al Coons]

Thanks Sean,

I input the data and create a linear model by ANALYZE/CURVE FIT/Linear, but it does not go through the origin.  Looked everywhere for a setting to force that.  What am I missing?

Al

--
--
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]

How about a screenshot of your data?

Below has (0,0) as one of the points.

​

---

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."

[Sean Bird]

Not the first choice 1:Linear, but the last choice B: Proportional.

Do you know that you don't need to retype the data, that is already in a list and spreadsheet? Menu > Data > Column Options (or click on the title of the list in Table view and choose "Link from List."

- Sean Bird

[Sean Bird]

Attached is the screen shot for the linear curve fit forced through the origin, a.k.a. a proportional equation, y = a*x

[Csaba Tizedes]

This "Proportional" option is a new thing for me.

If you want to fit y=a×x equation, simply calculate a=SUM(xi×yi)/SUM(xi^2). Where i=1, 2, ..., number of data points.

Csaba

[Thomas Hu]

I don't know if this helps, but you can create a line with the two points (0,0) and (xbar, ybar). Of course this is not as straightforward as using stat - regression menu, which unfortunately does not have the proportion option though available in vernier page.

Another option is use matrix approach with (AtA)^-1 * Atk formula to treat the general problem (linear in the unknowns), but this would not work if students don't know matrices.

[Csaba Tizedes]

Just read one comment above: "If you want to fit y=a×x equation, simply calculate a=SUM(xi×yi)/SUM(xi^2). Where i=1, 2, ..., number of data points."

Csaba

[Chris Engledowl]

Another option is to standardized both the dependent and independent variables. This will always result in passing through the origin because the point (0,0) will be the mean of both variables which is always on the regression line. 

Chris

--

--
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.

To view this discussion on the web visit https://groups.google.com/d/msgid/tinspire/f12c4d9d-dbca-49af-8f23-4b9031f7a37b%40googlegroups.com.

[Csaba Tizedes]

:) OK, but how you get the slope of the ORIGINAL data?

Csaba

2019. június 21., péntek 15:06:52 UTC+2 időpontban Chris Engledowl a következőt írta:

Another option is to standardized both the dependent and independent variables. This will always result in passing through the origin because the point (0,0) will be the mean of both variables which is always on the regression line. 

Chris

On Sun, Jun 9, 2019, 6:24 AM Csaba Tizedes <csaba....@gmail.com> wrote:

Just read one comment above: "If you want to fit y=a×x equation, simply calculate a=SUM(xi×yi)/SUM(xi^2). Where i=1, 2, ..., number of data points."

Csaba

2019. június 8., szombat 20:52:03 UTC+2 időpontban Thomas Hu a következőt írta:

I don't know if this helps, but you can create a line with the two points (0,0) and (xbar, ybar). Of course this is not as straightforward as using stat - regression menu, which unfortunately does not have the proportion option though available in vernier page.

Another option is use matrix approach with (AtA)^-1 * Atk formula to treat the general problem (linear in the unknowns), but this would not work if students don't know matrices.

--
--
To post to this group, send email to tins...@googlegroups.com

To unsubscribe send email to tins...@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 tins...@googlegroups.com.

[Chris Engledowl]

The standardized slope is the correlation. The slope of the original (i.e. original units) data won't go through the origin. Although I suppose you could ignore the intercept if it is not something you will interpret but then in order to plot it, the intercept would have to be put back in or the context would not remain intact. Why is it that you want it to go through the origin?

Chris

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.
To view this discussion on the web visit https://groups.google.com/d/msgid/tinspire/6234843e-7f59-469f-91e6-c75eae476138%40googlegroups.com.

[Sean Bird]

There is some data that is directly related or proportional, so forcing the regression to go through (0,0) would be satisfying. Also it could be helpful to compare the linear regression and the proportional regression.

Sean Bird

On Fri, Jun 21, 2019 at 11:40 AM Chris Engledowl <cengl...@gmail.com> wrote:

The standardized slope is the correlation. The slope of the original (i.e. original units) data won't go through the origin. Although I suppose you could ignore the intercept if it is not something you will interpret but then in order to plot it, the intercept would have to be put back in or the context would not remain intact. Why is it that you want it to go through the origin?

Chris

On Fri, Jun 21, 2019, 8:16 AM Csaba Tizedes <csaba....@gmail.com> wrote:

:) OK, but how you get the slope of the ORIGINAL data?

Csaba

[Chris Engledowl]

Doesn't the correlation indicate the proportionality? 

Chris

--

--
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.

To view this discussion on the web visit https://groups.google.com/d/msgid/tinspire/CAELrnPCPpXiPc1bqzozcqB8d9PYovbxEEtjhJUv8iqLL%2Bkzz3w%40mail.gmail.com.

[Kleofáš Slavný]

A short math backround

http://web.ist.utl.pt/~ist11038/compute/errtheory/,regression/regrthroughorigin.pdf

út 25. 6. 2019 v 3:28 odesílatel Chris Engledowl <cengl...@gmail.com> napsal:

To view this discussion on the web visit https://groups.google.com/d/msgid/tinspire/CA%2B0SBn5twvJj4SraUhr7sXmBsx__8H%3D1bT_OLh0bpamnX2DqOw%40mail.gmail.com.

[Sean Bird]

Thank so much for the link to the article. It was quite interesting.

...

- Sean Bird

On Sat, Jul 13, 2019 at 1:43 PM Kleofáš Slavný <k.sl...@gmail.com> wrote:

A short math background

http://web.ist.utl.pt/~ist11038/compute/errtheory/,regression/regrthroughorigin.pdf

To view this discussion on the web visit https://groups.google.com/d/msgid/tinspire/CAHqHNSSa4pDgXoAKb36P%3DqbZxB%3DiD54y%3DZDsOPHZx1dOMC%3D%2B8w%40mail.gmail.com.