Extract numbers from char type

how can I extract the numbers from a character array value of '4.64 km' and return them as 4.64?

Answer 1

[Star Strider]

Probably the easiest way —

s = '4.64 km';
c = regexp(s, '\d*\.\d*|\d*', 'match')

c = 1×1 cell array
    {'4.64'}

n = str2double(c)

n = 4.6400

EDIT — (18 Oct 2022 at 19:38)

In the event any of the numbers are negative —

s = '4.64 -42 2 -3.14 km';
c = regexp(s, '(\-\d*|\d*)\.\d*|(\-\d*|\d*)', 'match')

c = 1×4 cell array
    {'4.64'}    {'-42'}    {'2'}    {'-3.14'}

n = str2double(c)

n = 1×4

4.6400 -42.0000 2.0000 -3.1400

.

[Walter Roberson]

-?(\d+\.\d*¦\.\d+)

[Star Strider]

Noted!

I had forgotten that syntax and couldn’t find it in the documentation.

[Walter Roberson]

darn I forgot the case of no period

[Star Strider]

I saw that and wondered, however I did not change my answer, and just learned from the important parts of your syntax.

Answer 2

[Les Beckham]

Or, in one line of code:

s = '4.64 km';
num = sscanf(s, '%f') % num will be a double

num = 4.6400

[Walter Roberson]

Yes this works well when the number is at the beginning. You can also put literal text into the format if you expect an exact match.

The suggestions to use regexp to extract the text of the number work well when you need flexibility.

In cases where you have tables of text see textscan() which can be used on a character vector that has embedded newlines.

Answer 3

[David Hill]

a='4.65 km';
r=regexp(a,'\d*[.]*\d*','match');
str2num(r{1})

ans = 4.6500