0

我正在运行聚类分析,并希望在直方图中绘制一些结果。为了将一个集群与另一个集群进行比较,我发现为每个直方图使用相同的坐标系和相同的 bin 宽度很有用。对于情节范围,我已经得到了提示:

PlotRange/.Options[Histogram[DeleteCases[Flatten[data], {}], 
           PlotRange -> Automatic], PlotRange]

bin 宽度是否有类似的方法?使用我的方法,您会看到对于集群 3,bin 宽度很大,因为最低条超出了坐标系。

例子:

data = {{-1.2056`, -1.46192`, -1.30053`, -2.52879`, -0.99636`, {}},
{-1.73904`, -1.164`, -1.83398`, -0.97505`, {}, {}, -0.503256`,
-0.63802`}, {-0.785963`, -0.711821`, -0.820439`, -1.8699`, -3.9659`,
-1.4456`, -1.67021`, -1.42009`, -2.5644`, -1.45002`, -1.27806`,
-1.66529`, -1.67073`, -3.31102`, -3.38638`}}

Table[
 Histogram[
  DeleteCases[data[[i]], {}], 
  3*Abs[Differences[
      Last[PlotRange /. 
        Options[Histogram[DeleteCases[Flatten[data], {}], 
          PlotRange -> Automatic], PlotRange]]]]/
    Length[DeleteCases[Flatten[data], {}]],
  PlotLabel -> 
   Text[Style["Cluster " <> ToString[i], TextAlignment -> Center]],
  Frame -> {{True, None}, {True, None}},
  FrameLabel -> {{"###", None}, {"###", None}},
  PlotRange -> 
   Part[PlotRange /. 
     Options[Histogram[DeleteCases[Flatten[data], {}], 
       PlotRange -> Automatic], PlotRange], 2]
  ], {i, Length[data]}]
4

0 回答 0