Which tool is best for data plotting for scientific documents, journals, etc. In MATLAB, I did some plots with very low step size like the plot in the attachment, If I change different features like *, ^, v instead of '-' I am getting only thick lines (because of very low time steps). If I print in black and white printer, we cannot find the difference. How to overcome this issue. Can anyone suggest me a plotting tool to show the difference in black and white print out (we have to keep it in mind that we have to use very low time steps like 0.0005 also).
Although the calculation be done with time step of 0.0005 s, the results are presented in time steps of 10 s, as can be seen in the attached plot. I recommend reduce the number of points to be plotted and then utilize dots, traces etc. of different colours.
If you don't want to decrease the sampling, then you can play with intensity instead of colors. Make thicker lines and put different levels of white in the different curves.
If you are using Matlab and want to use markers then there are some suggestions on the Matlab Central site (http://uk.mathworks.com/matlabcentral/newsreader/view_thread/147920). Govind has posted a short code to demonstrate the approach.
In the "old days" we used to use dashed, dotted, dot-dashed, etc lines to show the differences!
I would suggest GnuPlot -- a free tool available both under UNIX and Windows (maybe other systems as well). A highly-customizable software, you can easily edit the output EPS if desired. Lots of manuals and samples available in the Web.
You could also plot the data twice once with the fine resolution as just a line and once as just a marker every so often....
If your data is contained in the parameters x and y
plot(x,y,'b-')
hold on
plot(x(1:10:end),y(1:10:end),'LineStyle','none','Marker','s','Color','blue')
hold off
These two lines before your sequence of plot() will set all plots lines to black, and will cycle automatically over different line styles for each plot:
set(0,'DefaultAxesColorOrder',[0 0 0]);
set(0,'DefaultAxesLineStyleOrder',{'+-','+','o-','o','x-','x','d-','d','^','v','>','
Use either EXCELL or Number Cruncher Statistical System (NCSS) SOFTWARES, THEY DISPLAY COLOURS .
you mean that your step size is very small, and you have thousands of data points in the figure, such that the * or ^ are not able to show the different line style? It is not recommended to have too much data points in a plot. It is suggested to use the interpolation method to make the figure more observable. for example, you can use
yy = interp1(x,y,0:2:100), to make the line with only 50 points.
you can interpolate each line to make the line clear.
If you use the interp1 route, I think it is worth noting that by default interp1 uses linear interpolation. yy = interp1(x,y,xq,method)
where
x and y are your data points
xq is where you would like to interpolate
and method can be the following (Assuming 2014a or higher...)
'linear' - linear interpolation
'nearest' - rounds to the nearest y value
'pchip' - fits a cubic polynomial between values
'spline' - fits a cubic spline using the not-a-knot conditions
Thank you for all the Researchers for your response on my questions (posts).
- Badges
- Science topic
- Similar topics
- Mathematical Sciences
- Graphs