My "calculated" polynomial is not the same as the Trend line on the graph. The one on the graph seems to be correct.
Basically, I am using the LINEST function to create a polynomial equation inorder to extrapolate data. And also with a given equation, and a given y, find x.
I have a table something like this (much bigger in reality)
1 5
2 10
5 15
7 #N/A
9 20
12 25
I need to interpret the above data with the linest function, but for a variety of reasons i cant delete the error square. How can i get the LINEST to ignore the #N/A and the 7? otherwise it returns an error.
I am using Excel 2007. I have two sets of data. y-series is 58.61, 58.66, 58.71 and 58.76. x-series is 0.8313, 0.8309, 0.8305 and 0.8301.
I am trying to fit a quadratic function to this data. When I plot a curve and fit a trendline I get the coefficients for the equation ax^2+bx+c
a=1.3939e-11
b=-125
c=162.5225
The plotted line seems to fit the data fairly well.
However, I want to use the linest function and I am using the formula:
=INDEX(LINEST(B2:B5,A2:A5^{1,2}),1,1)
I change the index column number as appropriate to get the coefficients:
a=-75.23775
b=0
c=110.603763
The coefficients from linest seem to be way out. how I can get the output from linest to be the same as the trendline?
I'm trying to implement the linest formula in a programming language for my coursework.
I've looked on excel help but it only explains on how the function selects data.
how to the values are calculated and the steps?
I am calculating variables in Vba and want to find their co=effiicients in the for y=a +bx +cx^2. This works great if I output my variables to excel and then pick up this range however as I am generating the variables in vba and have an array of variables I would rather not have to keep spitting them out to excel then picking them up again. The three important lines in my code are:
c(nFix - j) = WorksheetFunction.Index(WorksheetFunction.LinEst(inpv, Application.Power(inps, Array(1, 2))), 1)
b(nFix - j) = WorksheetFunction.Index(WorksheetFunction.LinEst(Range(Vc), Application.Power(Range(SPd), Array(1, 2))), 1, 2)
a(nFix - j) = WorksheetFunction.Index(WorksheetFunction.LinEst(Range(Vc), Application.Power(Range(SPd), Array(1, 2))), 1, 3)
As you can see for c I am referring to inpv and inps these are arrays. I have tried using inpv(), etc but no luck
b and a work great, here I am reffering to ranges Vc and SPd. As I am going through this a lot of times it seems rather inelegant not to be able to do the whole thing in vba - i would rather not use the dump to spreadsheet then pick up again approach!
is there a way to use LINEST in a way in which the x values are unknown and the y values are known? The opposite of how the function usually runs...?
View 2 Replies View Relatedhow to get my LINEST function to work properly. Now, what I'd like to do is have it reduce or increase the order of the polynomial until R^2 equals 1.0, or the polynomial reaches 6th order, whichever comes first.
Can I do this with a bunch of nested IF ELSEIF statements, or is there a cleaner way to do this?
I am trying to obtain the 3 coeffs from a polynomial trendline equation.
The equation produced on the chart is
y = 3.1123x2 + 135.34x + 124.9 from the below info
X Y
1.0 250
2.5 500
5.0 890
7.5 1290
10.0 1800
But when I enter this array formula
{=LINEST(B2:B6,A2:A6^{1,2})}
I get different results.
-48.98,375.89,214.98
I can do least squares fitting utilizing LINEST and a worksheet. (See attachment.)
I would like to do the same calculations entirely within VBA.
I can specify the values as arrays within VBA :
Sub linest2VBA()
'
' linest2VBA Macro
'
' Keyboard Shortcut: Ctrl+Shift+A
'
Dim dXvalSqd(4), dXval(4), dYval(4), dFirstRegStat As Double
Dim i As Integer
Second Question:
Is there a way to get the remaining regression statistics within VBA by referencing only the three arrays: dXvalSqd, dXval, and dYval?
I need to know how to selec no-adjacent columns for the known_x's in the linest function:
Example:
'- A B C D E
1) 80 63 9 3 26
2) 80 63 9 3 21
3) 74 65 7 2 30
4) 74 65 7 2 20
5) 63 65 9 3 16
=LINEST(A1:A5,KNOW_X'S,TRUE,TRUE)
Is there any way to select for instance: column B and E as known_x's?
I heard about something where we can select total area and then choose which columns should be considered via "{column1, colum2}"..
That is all is missing for me to finish my project!!! Pls your help!!!
I have built a user defined function called: portAlpha.
Code:
Function portAlpha(assetCode As String, indexCode As String) As Double
Dim assetCol As Long
Dim indexCol As Long
Dim assetData As Range
Dim indexData As Range
With Sheets("Data")
[Code] ......
I am struggling to get this to work though. Everytime I step over this piece of code:
Code:
portAlpha = Application.WorksheetFunction.LinEst(Range("M4:M63"), Range("N4:N63"), 1, 1)
I get a "Type Mismatch Error". All the values in the range are numbers, and LINEST should return a double, so I am puzzled as to why there is a mismatch...
I have hardcoded the range into the formula just to try and get it to work. Eventually this will be two variables and will look like this:
Code:
portAlpha = Application.WorksheetFunction.Index(Application.WorksheetFunction.LinEst(assetData, indexData, 1, 1),1, 1)
I am trying to use the LinEst function in a program. There are some good examples on the web but they all refer to a range in the format “A1:A45”. Is there a way to refer to a column in an existing array instead?
For example (thanks to Shg for this code): ....
I have set up regression in excel using the linest array function which automaticly updates whenever the data it is references changes. Currently the formula in K3 =
=LINEST(K8:K50,M8:N50,TRUE,TRUE) .
Only the Y range in column K changes and then the formula automaticly recalculates the regression parameters. However sometimes the range of valid data changes with each new update so currently K8 to K50 contains valid values and so the formula works. But if the range of valid data now ends at say K45, and the remaining rows (K46- K450) contains #DIV/0!. In this case the formula does not work and I need to go in manualy into K3 and reselect the range to get it to work (in the baove case to =LINEST(K8:K45,M8:N45,TRUE,TRUE)
Is there anyway I can get the formula to update the range (the number of rows to use) in the calculation automaticly so I don't have to go into the formula in K3 and change the range manually everytime the data changes? Basicly I like to use all rows in column K (from K8 to K50) where there is a values and ignore cells with #DIV/0!
I'm guessing the easiest way to do this would be to set up a dynamic range of some sort that feeds into the linest array function so that the end row reference for columns K and M & N changes depending on where the first #DIV/0! occurs in column K.
I have instances where my Y variables sometimes contain a zero in the data and i need a formula to skip over that data point in the LINEST calculation.
Y variables are in Row 1, Columns A:E
X variables are in Row 2, Columns A:E
The following formula is returning a #VALUE! error: .....
I'm using Excel 2003 and I'm trying to make my LINEST function and the source cells for a scatter graph use the cell range specified in another cell in the document. I figured out that it was simple to do the following:
I250 contains: 10, shows: 10
I251 contains: ="I"&I250, shows I10
But I can't figure out how to do this in a larger function. I've tried a few things but none of them work. Here is an example where I want it to do a LINEST with y values in cells Ja - Jb and x values in cells Ia - Ib (where a and b are integers specified in cells I250 and J250 respectively):