DScience:3D representation of data

Table of contents

30% complete

3D plots use 3D computer graphics to represent scientific data. Data are usually defined as $Z=f(X,Y)$, where $(X,Y,Z)$ values by a function or actual data. There are a number of possible types of 3D plots. You can use the following Java classes to plot data in 3D:  jhplot.HPlot3D,  jhplot.HPlotXYZ,  jhplot.HPlot3DP,  jhplot.HVisAd,  jhplot.HPlotMX. See additional details in DMelt:Plots/5_Plots_in_3D.

Bar plots or histograms

A bar plot draws a three-dimensional bar chart, where each element in Z corresponds to one bar. To make such plots from a 3D histogram use the Java class  jhplot.H2D. Here is the Python code with an example:

from jhplot  import HPlot3D,SHPlot3D,H2D
from java.util import Random


# build a standard canvas
c1 = HPlot3D("Canvas",600,400)

# build a singleton
# c1=SHPlot3D.getCanvas()

c1.setGTitle("Global title")
c1.setNameX("Xaxis")
c1.setNameY("Yaxis")
c1.visible(1)

h1 = H2D("My 2D Test",20,-3.0, 3.0, 20, -3.0, 3.0)
rand = Random();
for i in range(500):
               h1.fill(rand.nextGaussian(),rand.nextGaussian())
c1.draw(h1);

# export to some image (png,eps,pdf,jpeg...)
# c1.export(Editor.DocMasterName()+".png")

Surface plots

Here is an example of surface plots:

from jhplot import *
from java.awt import Color
from java.util import Random

c1=HPlotXYZ("test",600,400)
c1.setGTitle("3D function")
c1.visible()
c1.setTickDecimalAll(1)
hh=H2D("Histogram",20,-2,2,20,-20,20)
rand = Random()
for i in range(10000):
      hh.fill(rand.nextGaussian(),5*rand.nextGaussian())

c1.setLegendBar()
c1.addAsSurface(hh)
c1.animate()
c1.update()

Scattered data

Here is an example that use two data points (blue and red):

from jhplot  import *
from java.util import Random
from java.awt import Color

c1 = HPlot3D("Canvas",500,500)
c1.setGTitle("Interactive 3D galaxy") 
c1.setRange(-10,10,-10,10,-10,10)
c1.setNameX("X")
c1.setNameY("Y")
c1.visible(1)


# create P2D objects in 3D 
p1= P2D("Galaxy")
p1.setSymbolSize(2);
p1.setSymbolColor(Color.blue);

rand = Random()
for i in range(5000):
      x=3*rand.nextGaussian()
      y=3*rand.nextGaussian()
      z=0.4*rand.nextGaussian()
      p1.add(x,y,z)
c1.draw(p1)

h2=P2D("Core")
h2.setSymbolSize(2) 
h2.setSymbolColor(Color.yellow)
for i in range(5000):
      x=0.9*rand.nextGaussian()
      y=0.9*rand.nextGaussian()
      z=0.8*rand.nextGaussian()
      h2.add(x,y,z)
c1.draw(h2)

Mixed plots

One can overlay scattered data with surface plots, bar plots in arbitrary way. Typically, you need to use  jhplot.HPlot3D canvas.