multiple digit number in to individual digits

i want to change the number 1123 in 1 1 2 3, want to split combine number into into individual numbers

Answer 1

[Azzi Abdelmalek]

a=1234
b=str2double(regexp(num2str(a),'\d','match'))

[naveen selvan]

Hi, I have a follow-up question to ask. if you have a postfix expression like 123*+ then how do you split them into individual characters. For example, a = 123*+ then i need to get b = 1 2 3 * + pls help!

[Walter Roberson]

a = '123*+'
for K = 1 : length(a)
  fprintf('character #%d of "%s" is "%c"\n', K, a, a(K));
end

[Adam Danz]

For large values such as a=11122333345555566 this will not work since num2str will convert the value to '1.112233334555557e+16'. Otherwise nice solution.

Answer 2

[John D'Errico]

N = 1123;
Ndigits = dec2base(N,10) - '0'
Ndigits =
     1     1     2     3

[Adam Danz]

Great solution. Note that N must be smaller than flintmax.

[Michael Ramage]

How to do this with a floating point number?

[John D'Errico]

@Michael Ramage

Not difficult with a floating point number, but remember that a float is NOT an exact decimal representation of that number. But...

X = 1.2345;
dec2base(X*10000,10)

ans = '12345'

or

dec2base(X*10000,10) - '0'

ans = 1×5

1 2 3 4 5

You can even fuss around and get the decimal point in there if you want, but if you want that, then sprintf is arguably a better choice.

Answer 3

[Jan]

For getting the digits, a conversion to a string is an indirection. Staying at numerical values is usually faster:

N = 1123;
m = floor(log10(N));  % [EDITED] Thanks Stephen
D = mod(floor(N ./ 10 .^ (m:-1:0)), 10);

[Stephen23]

Almost right, but not quite:

 >> N = 10;
 >> m = ceil(log10(N));
 >> D = mod(floor(N ./ 10 .^ (m-1:-1:0)), 10)
 D =
     0

It ate my ten!!!!

Using floor not ceil works:

 >> N = 10;
 >> m = floor(log10(N));
 >> D = mod(floor(N ./ 10 .^ (m:-1:0)), 10)
 D =
      1     0

[Jan]

Thanks, Stephen. I've fixed it. Perhaps I need some holidays.

[Adam Danz]

Quite robust!

[Jay Doshi]

If the number is 0006, and i want all four numbers. What can I do? Because this method just gives me d = 6.

[Stephen23]

@Jay Doshi:

N = 6;
m = 3; % order
d = mod(floor(N ./ 10 .^ (m:-1:0)), 10)

d = 1×4

0 0 0 6

Answer 4

[Ramon Villamangca]

a simple single line solution:

>> num = 12345042117;
>> arrayfun(@(x) mod(floor(num/10^x),10),floor(log10(num)):-1:0)

ans =

     1     2     3     4     5     0     4     2     1     1     7

[Jyahway Dong]

62229893423380308135336276614282806444486645238749

apparently this number above doesn't work for the solutions proposed above, is it simply too long for MatLab to handle?

[Stephen23]

@Jyahway Dong: read about floating-point numbers and their properties:

https://www.mathworks.com/help/matlab/matlab_prog/floating-point-numbers.html

https://www.itu.dk/~sestoft/bachelor/IEEE754_article.pdf

For a number with that precision you will need to use the symbolic toolbox, or VPI class, or something similar:

https://www.mathworks.com/help/symbolic/index.html

https://www.mathworks.com/matlabcentral/fileexchange/22725-variable-precision-integer-arithmetic

[Ramon Villamangca]

@Jyahway Dong: If the digits are that long, you'll probably input it as a char string, anyway. That means the solution would even be much simpler:

num = '62229893423380308135336276614282806444486645238749';
num - '0'

ans = 1×50

6 2 2 2 9 8 9 3 4 2 3 3 8 0 3 0 8 1 3 5 3 3 6 2 7 6 6 1 4 2