asked on

Slope

Hello,

I am trying to find the linear regression of weight gain over time via MS SQL Server. I found a function for slope and im trying to apply it. I am looking for a conformation on this equation and if im applying it correctly.

Slope = ( COUNT(*)*SUM(x*y) -SUM(x)*SUM(y) ) / (COUNT(*)*SUM(x^2)-SUM(x)^2)

The site i found it said x is date and y is weight. I figured (correct me if im wrong) that x would be the age of the bird in days so i came up with the following query....

SELECT
      (COUNT(*)*SUM(Cast(DateDiff(dd, [Hatch Date], [Date]) As Float)*Weight) -SUM(Cast(DateDiff(dd, [Hatch Date], [Date]) As Float))*SUM(Weight) ) / (COUNT(*)*SUM(Cast(DateDiff(dd, [Hatch Date], [Date]) As Float)^2)-SUM(Cast(DateDiff(dd, [Hatch Date], [Date]) As Float))^2)
FROM
      WingbandFlock As wf
INNER JOIN
      Weight As w
ON
      wf.[Bird ID] = w.[Bird ID]

In concept, am i correct? I am still running into an SQL error i have not solved, if you know that you can post the answer in my related question for more points :)

https://www.experts-exchange.com/questions/22077814/Linear-Regression-Slope.html?anchorAnswerId=18045854#a18045854

Thanks!

-Navicerts

ASKER CERTIFIED SOLUTION

ozo

membership

This solution is only available to members.

To access this solution, you must be a member of Experts Exchange.

Start Free Trial

Navicerts

ASKER

Well shoot, that explains some things. Working now! Sorry i *thought* it was a math question, but you answered an SQL Server question :) I graphed it and the answer checks out too.

Thanks again!

SELECT
      (Count(*)*SUM(Cast([Feed Consumed] As float)*([Weight Gain]*453.6)) -SUM(Cast([Feed Consumed] As float))*SUM(([Weight Gain]*453.6)) ) / (Count(*)*SUM(power(Cast([Feed Consumed] As float),2))-power(SUM(Cast([Feed Consumed] As float)), 2))
FROM
      WingbandFlock As wf
INNER JOIN
      [Feed Consumed] As fc
ON
      wf.[Bird ID] = fc.[Bird ID]
INNER JOIN
      [Keepers Report] As kr
ON
      wf.[Bird ID] = kr.[Bird ID]