Hello
I would to have an udf that presentes the evolution of the daily radiation (by collecting the x and y (of a curve of an article for validation) from the "GET DATA"). At First, I drew my curve on excel, I obtained my trend curve and then the equivalent equation in the form of a polynomial of a certain degree ( I, who controls the degree to have a trend line close to the mine, genarally I had a degree that is equal to or greater than 4).
The problem is that when I drew the equation on matlab to confirm the tracing of the curve, I would have a different curve to the one I have in excel. What will be the resolution if someone of you has already had the same problem?
Thank you
cordially
Hi Lamaamar Ibtissam,
Since no plot is provided, I'm not exactly sure of the problem that you described. But I generally figured out the issue. Consider the figure above.
The polynomial trend (dotted line) does not exactly match the data series (solid line) plotted in MS Excel, though it can fit all the data points perfectly (R2 = 1).
The reason is because the Excel uses the spline interpolation technique to display a smooth curve through the different data points. The spline chart may or may not describe the true behavior of the data.
Since you have access to the MATLAB® software, you can try to use the built-in Curve Fitting Toolbox. You should also try to obtain as many 'useful' data points as possible.
Dear Lamamar Ibtissam,
Also i suggest you to take a look at links and attached file on subject.
How To Create A Linear Standard Curve In Excel - Top Tip Bio
https://toptipbio.com/linear-standard-curve-excel/
Using Excel to find best-fit curves - SLU Math
math.slu.edu/~may/ExcelCalculus/sec-1-5-IntroBestFitCurves.html
How to Make a Graph for Density in Excel With Microsoft | Techwalla ...
https://www.techwalla.com › Around The Home › Productivity
Graphing with Excel
web.iitd.ac.in/~nkurur/2011-12/Isem/cyp501/ExcelGraph.pdf
Best regards
Increase your data points as much as possible since less data points my be fitted with multiple functions, which of course, behave in different way for in-between data points
Thanks you all for your answers
Here is the equation I got from excel, but when I retrace it on matlab, the results were different .
I canot solve the problem
Hi Lamaamar Ibtissam,
Since there are no trendline and R-squared value, I do not know how close the data are to the Excel's fitted polynomial trendline.
As you can see from the figure above, the 3-point data (blue solid line) are fitted to the polynomial trendline (blue dotted line), y = − 0.0067x2 + 0.41x − 5.8833.
This parabolic quadratic equation is then used to generate the 9-point data (orange solid line). However, there is an obvious discrepancy between the blue solid line and the orange solid line.
One explanation for this discrepancy is that the decimals of the coefficients of the quadratic equation displayed (due to rounding to nearest significant figures) for the blue solid line can cause some loss of precision.
- Badges
- Science topic