# Regression Analysis without Excel or other program

#### ddap210

##### New Member
Hi All,

I was wondering if it is possible to manually calculate the X Variable Coefficient that is the output when the Excel LINEST function is used AND the const parameter is set to TRUE? I'm working inside an application that cannot interface with another program/application such as Excel or other. The application I'm working in can do basic SLOPE, Yintercept and Correlation functions but no functions are available to match the LINESTR with const = TRUE.

If const is set to TRUE then the coefficient = 0.000133633
If set to FALSE then the coefficient = 0.000137441

X Y
0 0
2 0.000208
4 0.000578
6 0.000793

Any help would be appreciated, even if the answer is no this cannot be done.

Thank you!

#### Miner

##### TS Contributor
When you set the const to TRUE, the equation fit is y = b + mx. When const is set to FALSE, b is forced to zero and the equation that is fit is y = mx. Coefficient is m.

#### Dason

If const is set to TRUE then the coefficient = 0.000133633
If set to FALSE then the coefficient = 0.000137441
Do you have those flipped around?

#### ddap210

##### New Member
Argh...yes I have it flipped in my initial post...my apologies. Thank you checking!

#### Dason

So if you had that flipped around are you still looking to match the results from when const = TRUE or do you want const = FALSE?

Basically do you want to find m when fitting:

y = mx + b

or

y = mx

Also - what application are you using?

#### ddap210

##### New Member
I would like to calculate FALSE. The program is a LIMS system - Laboratory Information Management System which has it's own language (code) that does not have a function calculate regression analysis where constant = 0. The data set has 1 independent variable and 4 dependent variables, if I understand this correctly.

I have attached a screen shot that may clear things up a bit.

#### Attachments

• 26.1 KB Views: 3