Splitting number from its units

HI, I have a cell array with following values:

'1mcg/kg'
'1mcg/kg'
'1mcg/kg'
'0.7mcg/kg/hr'
'0.7mcg/kg/hr'
'0.5mcg/kg/hr'
'0.5mcg/kg/hr'
'0.5mcg/kg/hr'

How do i split this into numbers and units? I need the output in two cell arrays something like:

'1' 'mcg/kg'
'1' 'mcg/kg'
'0.7' 'mcg/kg/hr'
'0.5' 'mcg/kg/hr'

Answer 1

[per isakson]

Try

>> cac = regexp( '0.7mcg/kg/hr', '([\d\.]+)|(\D+)', 'tokens' );
>> cac{:}
ans = 
    '0.7'
ans = 
    'mcg/kg/hr'
>> cac = regexp( '0.7mcg/kg/hr', '([\d\.]+)|(\D+)', 'match' );
>> cac{:}
ans =
0.7
ans =
mcg/kg/hr
>>

and read expression — Regular expression

Note: Numbers in the unit string will cause problems

&nbsp

This returns leading digits and dots as the value and the rest of the string as the unit.

>> cac = regexp( '0.7mcg2/kg/hr', '\<([\d\.]+)(.+)\>', 'tokens' );
>> cac{:}
ans = 
    '0.7'    'mcg2/kg/hr'
>>

Whatever mcg2 stands for.

And as Azzi shows regexp takes a cell array of strings in place of a single string.

Answer 2

[Image Analyst]

As an alternative, if you don't understand the cryptic regexp commands, then you can use a simple "for" loop:

ca = {'1mcg/kg'
'1mcg/kg'
'1mcg/kg'
'0.7mcg/kg/hr'
'0.7mcg/kg/hr'
'0.5mcg/kg/hr'
'0.5mcg/kg/hr'
'0.5mcg/kg/hr'}

for k = 1 : length(ca)
  thisString = ca{k}; % Extract this string
  % Find where the units start.  They will start with a letter.
  firstNonNumberIndex = find(thisString >= 'A', 1, 'first');
  % Extract numbers and units into two cell arrays.
  numbers{k} = str2double(thisString(1:firstNonNumberIndex-1));
  units{k} = thisString(firstNonNumberIndex:end);
end
% Show them in the command window
celldisp(numbers);
celldisp(units);

I'd really recommend though that you don't use a cell array for the numbers and use it only for the units. Use a regular numerical double array for the numbers.

Answer 3

[Azzi Abdelmalek]

str={'1mcg/kg'
'1mcg/kg'
'1mcg/kg'
'0.7mcg/kg/hr'
'0.7mcg/kg/hr'
'0.5mcg/kg/hr'
'0.5mcg/kg/hr'
'0.5mcg/kg/hr'}
v1=regexp(str,'[\d\.]+','match','once')
v2=regexpi(str,'[a-z\/]+','match','once')