« View all new features in
Mathematica
9
◄
previous

next
►
New in
Mathematica
9
›
Builtin Integration with R
Seed Random Number Generation and Visualization
In[1]:=
X
Needs["RLink`"] InstallR[]
Generate random normally distributed numbers with a fixed seed.
In[2]:=
X
REvaluate["{ set.seed(123) rnorm(10,10) }"]
Out[2]=
Package this as a function.
In[3]:=
X
rnormFixedSeed = RFunction["function(n, m = 0, sdev = 1, seed = 1){ set.seed(seed) rnorm(n,mean = m, sd= sdev) }"];
Test the function.
In[4]:=
X
rnormFixedSeed[10]
Out[4]=
Since the seed is fixed, the same seed value always returns the same pseudorandom sequence.
In[5]:=
X
rnormFixedSeed[10]
Out[5]=
Generate a set of random numbers.
In[6]:=
X
(randR = rnormFixedSeed[10000]) // Short // AbsoluteTiming
Out[6]=
Test them for normality with
Mathematica
's
QuantilePlot
.
In[7]:=
X
QuantilePlot[randR]
Out[7]=
And with
EstimatedDistribution
.
In[8]:=
X
Clear[mu, sigma]; dist = EstimatedDistribution[randR, NormalDistribution[mu, sigma]]
Out[8]=
Visualize the result.
In[9]:=
X
Show[Histogram[randR, Automatic, "ProbabilityDensity"], Plot[PDF[dist, x], {x, 5, 5}, PlotStyle > Thick]]
Out[9]=
Use the dynamic interactivity of
Mathematica
for data analysis. The random numbers are generated in R but analyzed and visualized in
Mathematica
.
In[10]:=
X
Clear[mu, sigma, m, s]; Manipulate[ With[{randR = rnormFixedSeed[n, m, s]}, With[{dist = EstimatedDistribution[randR, NormalDistribution[mu, sigma]]}, Show[Histogram[randR, Automatic, "ProbabilityDensity"], Plot[PDF[dist, x], {x, m  5 s, m + 5 s}, PlotStyle > Thick]] ]], {{n, 1000, "Number of points"}, {100, 500, 1000, 2000, 5000, 10000, 50000}}, Grid[ { {Row[{Control[{{m, 0, "Mean"}, 0, 100}], Framed@Dynamic[m]}, " "]}, {Row[{Control[{{s, 1, "Standard deviation"}, 0.01, 10}], Framed@Dynamic[s]}, " " ]} }]]
Out[10]=