I used Graphpad prism to determine which order polynomial I should use for each data set, comparing best fit between two models at a time (ie linear and quartic, then quartic and cubic). I have now calculated my R values and slopes for various order polynomials. Hope this is clear and answers your question. The rest should only take a minute or so for each subject. The longest part of the process will be entering your data for each subject.
In this way you can put the formulas in the template as in the attached file which includes formulas down to row 100. Iv) save the template with a new file name for that subjectĪ further improvement would be to wrap the formulas for the polynomials and slope in an if statement so that results are only displayed if the y cells are not blank.
#EXCEL TRENDLINE POLYNOMIAL UPDATE#
The values for the coefficients will then automatically update to the new values Iii) You will then be asked if you want to replace the the existing x, y and Formula values, so press Yes for each. Ii) On the menu bar select Insert/Name/Create and tick Top Row then OK Copy the slope formula from cell K2 and paste into cells K3:Kn
Copy the polynomial formula from cell C2 and paste into cells C3:Cn (where n is the last row of data)ģ. enter the x and y data (if the x values are always the same and n is always 90 you could put the x values in the template).Ģ. You can then use the template to create a file for each subject. You can also do a quick visual check to show that the slope value passes through zero around the x values for the maximum and minimum of the trend line, and that the areas of negative slope are 'downhill' and positive slope are 'uphill' on the trend line.įor several large sets of data the best thing to do is keep one file as a sort of template.
0 kommentar(er)
