Please don't use this for a published plot I'm not responsible for you being stoned, tarred and feathered at a conference, or worse, denied tenure. Like I said, unfinished or not production. You might want to remove the DEBUG define from plot2table.pro, otherwise the output file is always "points.dat", and there is a lot of stuff printed to the terminal. There are a whole boatload of things I haven't implemented yet. When launching plot2table, you can give an image as a command-line argument, and it will be automatically placed in the Open dialog.Įrror bar input is still undocumented (since I haven't settled on a scheme). I've included a simple plot for tests, dimu_v2.gif. My "To Do" list is in notes.rtf, sorry if you can't read Mac rich-text files.īinaries for Windows, Mac, and Linux are written to plot2table.exe, plot2table.app, and plot2table, respectively. The configuration itself is stored in plot2table.pro. All other configurations are at your own risk (it was running on Scientific Linux several versions ago). Pixel perfect simple element server error icon for web design, apps. It has been built using Qt 5.2 on Mac OSX 10.9.5. Find the perfect server error stock photo, image, vector, illustration or 360 image. This project is still in an unfinished state. Here is a screenshot of it running on my Mac: There is currently preliminary support for error bars, but that is still WIP. The numerical precision is of course limited by the pixel size and image quality. The program then uses the defined pixel->x,y mapping to create a table. It works by clicking two points (one on each axis) and defining what the x,y values are at those points, then clicking on all the data points that are to be saved. Once you have it in csv you can just import it into excel and do your analysis that way, but I think doing it in python is by far the superior way to go.Plot2Table was designed to allow a user with a graphical plot to create a table of numerical values instead. In that case what I would do is crop out everything but the waveform, and then run it though a bitmap-to-csv converter such as this one I found from a cursory google search. If you're not comfortable writing your own code (a good engineer should be, though), there are tons of hits on google for "bmp to csv converter". Two nested for-loops can probably iterate the entire image in less than 10 seconds on a modern laptop. I'm sure there's a more sophisticated way to do it but frankly sometimes the "dumb solution" is the simplest. ![]() If you just want the x-y coordinates of the green pixels, then before you save your data out you can just iterate through and check each pixel's RGB value against a threshold value (in case your output program does aliasing). You might have to do a little conditioning to size the array the way you want it based on the tuple returned by img.size() Numpy.savetxt("pixeldata.csv", pixeldata, delimiter=",") ![]() Engauge Digitizer Engauge Digitizer, like DataThief, works on Linux, Mac and Windows. In this case, if the output distance is too small, the matched curved will go along the edge of the marker (big error if the marker is not too small). Just from a quick glance at the PIL and numpy documentation online, I'd imagine the following would be sufficient to dump the entire thing into excel: from PIL import Image One defect of DataThief is the extraction of a line with marker. csv that you can open with Excel, you can use the numpy library. You can use PIL, the Python Imaging Library, to import the image before doing whatever processing you want. I personally am a fan of the python programming language for things like this.
0 Comments
Leave a Reply. |