New in Wolfram
Mathematica
8: Nonparametric, Derived, and Formula Distributions
◄
previous

next
►
Core Algorithms
Use Nonparametric Distributions Like Any Other Distribution
The wordlength distributions of English parts of speech are estimated and the expected length of a noun is computed from a
HistogramDistribution
.
In[1]:=
X
partOfSpeech = WordData[All, "PartsOfSpeech"]; data = Table[ StringLength /@ i, {i, (WordData[All, #] & /@ partOfSpeech)}];
In[2]:=
X
Framed[Grid[ Partition[ Table[Plot[ Evaluate[PDF[HistogramDistribution[data[[i]]], x]], {x, 0, 30}, PlotLabel > Style[partOfSpeech[[i]], Bold, FontFamily > "Verdana"], PlotRange > All, Exclusions > None, PlotPoints > 500, Filling > Axis, PlotStyle > {Thick, Black}, FillingStyle > ColorData[14, 8], ImageSize > 175, Ticks > {Range[0, 30, 10], None}, Axes > {True, False}], {i, Length[partOfSpeech]}], 3]], RoundingRadius > 10, FrameMargins > 10, Background > GrayLevel[.95]]
Out[2]=
In[3]:=
X
Expectation[len, len \[Distributed] HistogramDistribution[data[[1]]]] // N
Out[3]=