replacing the NaN with specific values

Question

tzaloupas 2013-3-14

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/67159-replacing-the-nan-with-specific-values

评论： Khaing Zin Htwe 2016-5-10

采纳的回答： Andrei Bobrov

在 MATLAB Online 中打开

Dear all,

I have the following matrix

   A={'country'  'values'
       'E'    '1'  
    'E'    [NaN]
    'E'    [NaN]
    'E'    '2'  
    'E'    [NaN]
    'E'    [NaN]
    'E'    [NaN]
    'E'    [NaN]
    'E'    [NaN]
    'E'    '3'  
    'E'    [NaN]
    'E'    [NaN]
    'E'    [NaN]
    'E'    [NaN]
    'EE'    [NaN]
    'E'    '4'
     'I'    '1'  
    'I'    [NaN]
    'I'    [NaN]
    'I'    '2'  
    'I'    [NaN]
    'I'   [NaN]
    'I'    [NaN]
    'I'    [NaN]
    'I'    [NaN]
    'I'    '3'  
    'I'    [NaN]
    'I'    [NaN]
    'I'    [NaN]
    'I'   [NaN]
    'I'    [NaN]
    'I'    '4'
    'I'    [NaN]
    'I'    [NaN]
    'I'    [NaN]
    'I'   [NaN]
    'I'    [NaN]
    'I'    '5'
     'K'    '1'  
    'K'    [NaN]
    'K'    [NaN]
    'K'     '2'  
    'K'     [NaN]
    'K'    [NaN]
    'K'     [NaN]
    'K'     [NaN]
    'K'     [NaN]
    'K'    [NaN]   
   'K'    [NaN]
    'K'    '3'
    'K'     [NaN]
    'K'    [NaN]
    'K'    [NaN]
    'K'     '4'
    'K'    [NaN]
    'K'     [NaN]
    'K'     [NaN]
    'K'    [NaN]
    'K'     [NaN]
    'K'     '5'
};

The first column describes the countries and the second the values corresponding to each country.

For each country the initial value is always '1'. For each country I want to replace the NaNs with the value that is before them.

So for the first country I want to have

'E'    '1'  
  'E'   '1' 
  'E'    '1' 
  'E'    '2'  
  'E'   '2' 
  'E'    '2' 
  'E'    '2' 
  'E'    '2' 
  'E'    '2' 
  'E'    '3'  
  'E'    '3' 
  'E'    '3' 
  'E'    '3' 
  'E'    '3' 
  'EE'    '3' 
  'E'    '4'
    for country I I want to have
        'I'    '1'  
    'I'    '1' 
    'I'   '1' 
    'I'    '2'  
    'I'   '2' 
    'I'  '2' 
    'I'    '2' 
    'I'   '2' 
    'I'    '2' 
    'I'    '3'  
    'I'   '3'  
    'I'    '3'  
    'I'    '3'  
    'I'   '3'  
    'I'    '3'  
    'I'    '4'
    'I'    '4'
    'I'   '4'
    'I'    '4'
    'I'   '4'
    'I'   '4'
    'I'    '5'
    and so so..
    I am trying to come up with a code but I got stuck
    Is there any way of doing that
    My real vector contains 300 countries, so doing manual is quite tedious
    thanks in advance

0 个评论
显示 -2更早的评论隐藏 -2更早的评论

请先登录，再进行评论。

请先登录，再回答此问题。

Answer 1

Andrei Bobrov 2013-3-14

1
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/67159-replacing-the-nan-with-specific-values#answer_78607

编辑：Andrei Bobrov 2013-3-14

在 MATLAB Online 中打开

A1 = A(2:end,:);
t = ~cellfun(@isnan,A1(:,2));
k = A1(t,2);
A(2:end,2) = k(cumsum(t));

added on comment

A1 = A(2:end,2);
t = cellfun(@ischar,A1);
A1(t) = cellfun(@str2double,A1(t),'un',0);
t1 = ~cellfun(@isnan,A1);
k = A1(t1);
A(2:end,2) = k(cumsum(t1));

5 个评论
显示 3更早的评论隐藏 3更早的评论

tzaloupas 2013-3-14

在 MATLAB Online 中打开

Hi Andrei,

you are the best. It works fine

I forgot to mention that for some countries I have fixed values ; see for country 'z' and'T'

   A={'country'  'values'
   'Z'    [50]  
    'Z'    [50] 
    'Z'    [50] 
   'Z'   [50]  
    'Z'   [50] 
    'Z'    [50] 
    'Z'    [50] 
    'Z'    [50] 
    'Z'    [50] 
    'Z'    [50] 
    'Z'    [50] 
    'Z'    [50] 
    'Z'    [50] 
   'Z'    [50] 
    'Z'    [50] 
    'Z'    [50] 
     'E'    '1'  
    'E'    [NaN]
    'E'    [NaN]
    'E'    '2'  
    'E'    [NaN]
    'E'    [NaN]
    'E'    [NaN]
    'E'    [NaN]
    'E'    [NaN]
    'E'    '3'  
    'E'    [NaN]
    'E'    [NaN]
    'E'    [NaN]
    'E'    [NaN]
    'EE'    [NaN]
    'E'    '4'
    'T'    [5]  
    'T'    [5] 
    'T'    [5] 
   'T'    [5]  
    'T'   [5] 
    'T'    [5] 
    'T'    [5] 
    'T'    [5] 
     'I'    '1'  
    'I'    [NaN]
    'I'    [NaN]
    'I'    '2'  
    'I'    [NaN]
    'I'   [NaN]
    'I'    [NaN]
    'I'    [NaN]
    'I'    [NaN]
    'I'    '3'  
    'I'    [NaN]
    'I'    [NaN]
    'I'    [NaN]
    'I'   [NaN]
    'I'    [NaN]
    'I'    '4'
    'I'    [NaN]
    'I'    [NaN]
    'I'    [NaN]
    'I'   [NaN]
    'I'    [NaN]
    'I'    '5'
     'K'    '1'  
    'K'    [NaN]
    'K'    [NaN]
    'K'     '2'  
    'K'     [NaN]
    'K'    [NaN]
    'K'     [NaN]
    'K'     [NaN]
    'K'     [NaN]
    'K'    [NaN]   
   'K'    [NaN]
    'K'    '3'
    'K'     [NaN]
    'K'    [NaN]
    'K'    [NaN]
    'K'     '4'
    'K'    [NaN]
    'K'     [NaN]
    'K'     [NaN]
    'K'    [NaN]
    'K'     [NaN]
    'K'     '5'
};

In that case when I apply the code I get this message

 Error using cellfun
Non-scalar in Uniform output, at index 1,
output 1.
Set 'UniformOutput' to false.

Could you help me modify the code. I am so close! Please?

Best

tzaloupas 2013-3-14

PERFECT!

Khaing Zin Htwe 2016-5-10

To replace integer zeros to NaN vaules in Array ,how can I do it ,sir.I have 141x3084 array that include NaN values unfortunately. After reducing dimensions to give input to ANFIS classifier, i get the result as NaN,seriously. Please help me ,sir , Andrei Bobrov.

请先登录，再进行评论。