Data Analysis - I have a graph. How do I convert it to an Excel File so I can make Calculations?

This is a question many engineers have.  We will provide some guidance below using free web-based tools we have found useful.

1. Take a screenshot of the graph and save it to your computer.
2. Use an online digitizer tool. There are many of these.  I use WebPlotDigitizer a free service provided by Ankit Rohatgi.
3. Interpolate the digitized data into equal spaced intervals. This can be done using http://polynomialregression.drque.net/online.php a site developed by Andrew Que.  I use this site because it allows you to force the regression through a reference point by using a weighting column.  You can also use the built in regression functions in Excel.

Example using the Relative Power vs. Junction Temperature plot in the SST-10-UV datasheet.

This plot has three curves, we will digitize the red line for the 385/395 nm component.

Here I have uploaded the screenshot file.  The first step is to align the axis.  This software is intuitive so I will omit steps and just give the big picture.

Here I have clicked on my reference points and entered the values I want to use for the digitization operation. I prefer to use 0-1 scaling rather than percents.  

 After I click OK, I want to digitize the red line. This can be done by just manually clicking the curve but I want to demonstrate the color and pen options that can be used to reduce the number of clicks needed.

Click the blue box on the right. This will bring up the color picker tool.  Click on the color picker button, then the red line in the plot, then Done.  This will tell the pen tool to look for that particular shade of red.

Click on Pen and trace over the red curve.  The default pen width is usually OK but you can change it if needed. Click Run.

Here is the result after hitting Run.  This result is very clean and does not need any editing.  Sometimes you will need to add or delete points which can be done using the menu on the right.

Copy the points onto your clipboard - View Data -> Copy to Clipboard.

Since we didn't edit any points, the data is already sorted.  If you do edit points, you should sort the data in Excel after you paste it in. I am not going to cover how to import data into Excel, but here is the result after I added labels and a default xy plot. 

These data looks good but the spacing is not very useful.  For these data, you could use the Excel regression tools to generate new spacings but I want to demonstrate how to force the regression line to go exactly through the 25 C, 1.0 reference point. 

Open http://polynomialregression.drque.net/online.php.  I already know that these data have a cubic form so I will select that option. Below I have selected a cubic regression with weighting using the third column option.

Go back to Excel and add a weighting column to your data.  Add your reference point and give it a big number for the weight.  Set all the other weights to one.  This website does not recognize scientific notation so make sure your reference point weight is an integer when you copy the numbers.

Paste your data in the data entry box and click submit.  The results are shown below.  Copy from Function to the R-Squared value and paste into Excel.

Use the regression coefficients to generate your desired spacing.

Useful Links:

Microsoft Article on LINEST: https://support.microsoft.com/en-us/office/linest-function-84d7d0d9-6e50-4101-977a-fa7abf772b6d

How to use the LINEST function for polynomial LMS fits: https://people.stfx.ca/bliengme/exceltips/polynomial.htm

Confidence intervals, multiple variable regressions: http://www.tushar-mehta.com/publish_train/data_analysis/16.htm

Using sum of squares minimums and the Excel solver to regress arbitrary functions: https://jkp-ads.com/articles/leastsquares.asp

Using Python instead of Excel: https://luminusdevices.zendesk.com/hc/en-us/articles/4418755468941-Data-Analysis-Interpolating-and-Curve-Fitting-with-Python

--------------------------------------------------------------------------------------------------------------------

Luminus Website https://www.luminus.com/

Luminus Product Information (datasheets): go to the main page and select Products and then select the type of product you are interested in.

Luminus Design Support (ray files, calculators, ecosystem items: [power supplies, lenses, heatsinks]): go to the main page and select Resources & Tools and select the item you are interested in.

Luminus Product Information sorted by Applications:  go to the main page and select Applications.

Where to buy Samples of Luminus LEDs: https://www.luminus.com/contact/wheretobuy.

--------------------------------------------------------------------------------------------------------------------

Technical Support Contact Information: [e-mail]

Sales Support Contact Information: [e-mail]

Customer Service Support Contact Information: [e-mail]