茎叶图算法

时间:2017-01-31 20:38:51

标签: matlab

我正在尝试在MATLAB中实现一个词干和绘图算法用于教育目的。在发布我的代码之前,让我先介绍一下我的方法的步骤。让我们考虑一下我们有两位数字:

A=[20 12 13 21 56 13  16 17 22 23 24];

茎可以通过

给出
stems=fix(A/10)
stems =
     2     1     1     2     5     1     1     1     2     2     2

和叶子可以由

给出
leaf=fix(mod(A,10))

leaf =

     0     2     3     1     6     3     6     7     2     3     4

我所做的,就是对茎进行分类,并根据那种叶子进行分类:

[stems, index]=sort(stems,'ascend')
leaf=leaf(index)
stems =
     1     1     1     1     1     2     2     2     2     2     5
leaf =
     2     3     3     6     7     0     1     2     3     4     6

这是基本的想法:

    1. 计算stems
    2. 中每个数字的出现频率
    1. leaf
    2. 中取出许多元素

对每个词干重复此过程,在每个步骤中我缩短leaf数组。例如,对于stems = 1,我们有[5 1],所以我会有

leaf(1:5)
ans =
     2     3     3     6     7
leaf(1:5)=[]
leaf =
     0     1     2     3     4     6

stems = 2再次5次,所以再次:

leaf(1:5)
ans =
     0     1     2     3     4
leaf(1:5)=[]
leaf =
     6

现在stems = 5,我们有1片

leaf(1)
ans =
     6

为此,我使用了一个地图容器,并创建了以下代码:

function stem_leaf_plot(v)
if ~isnumeric(v)  %  check that  program will accept  array as a  integers
    error( 'Input V must be numeric'); 

end
stems=fix(v/10);
leaf=fix(rem(v,10));
[stems, index]=sort(stems,'ascend');
leaf=leaf(index);
string_stems=num2str(stems);
%%  count  occurence of each stem
MAP=containers.Map();
n=length(stems); % total element of  stems array
for  ii=1:n
    if isKey(MAP,string_stems(ii))
        MAP(string_stems(ii))= MAP(string_stems(ii))+1;
        else
         MAP(string_stems(ii))=1;
    end
end
MAP_count=length(MAP);

stem=num2str(cell2mat(keys(MAP)));
for jj=1:MAP_count
    frequency=(MAP(string_stems(jj)));
    fprintf('leafs of stem %d',stem(jj));
    disp(leaf(1:frequency));
    leaf(1:frequency)=[]; % delete   elements  step by step
end

end

但是,我的代码的结果是

stem_leaf_plot(A)
leafs of stem 32     2     3     3     6
leafs of stem 49     7     0     1     2     3     4     6

有什么问题?

1 个答案:

答案 0 :(得分:0)

@Adriaan的建议之后,我使用hist来计算频率,而不是容器。这是我更新的代码:

function stem_leaf_plot(v)
if ~isnumeric(v)  %  check that  program will accept  array as a  integers
    error( 'Input V must be numeric'); 

end
stems=fix(v/10);
leaf=fix(rem(v,10));
[stems, index]=sort(stems,'ascend');
leaf=leaf(index);
[a,b]=hist(stems,unique(stems));
n=length(a);
for ii=1:n
    fprintf('leaf of  stem  %d is ',b(ii));
    leaf(1:a(ii))
    leaf(1:a(ii))=[];

end


       >> A=[20 12 13 21 56 13  16 17 22 23 24];
>> stem_leaf_plot(A)
leaf of  stem  1 is 
ans =

     2     3     3     6     7

leaf of  stem  2 is 
ans =

     0     1     2     3     4

leaf of  stem  5 is 
ans =

     6