Put the separator every thousands

T=100000000;
T1 = regexprep(string(T),'(\d+)(\d{3})$',"$1'$2")

T1 = "100000'000"

I would like to have the thousands separator for every 1000

the correct result would be:
1'000'000'000

[Stephen23]

"the correct result would be: 1'000'000'000"

Why should the "correct result" be one billion when the input value is only one hundred million?

100000000  % your input value
1000000000 % your "correct result" with quotes removed

Your "correct result" is ten times larger than the input value: is that intentional or is it ... incorrect ?

[shamal]

Yes, I counted the zeros wrong :DI to communicate 1000000000 become should 1'000'000'000

Answer 1

[Stephen23]

Here are some more interesting testcases (with both one hundred million as well as one billion):

S = ["100000000";"1000000000";"123456789.123456789";"-123456789";"0";pi;"1e23456"]

S = 7×1 string array
    "100000000"
    "1000000000"
    "123456789.123456789"
    "-123456789"
    "0"
    "3.1416"
    "1e23456"

S = regexprep(S,'(?<!(\.|[eE][-+]?)\d*)\d{1,3}(?=(\d{3})+(e|E|\.|\>))', '$&,')

S = 7×1 string array
    "100,000,000"
    "1,000,000,000"
    "123,456,789.123456789"
    "-123,456,789"
    "0"
    "3.1416"
    "1e23456"

[shamal]

Good 👍👍

Answer 2

[Image Analyst]

I use the attached function I wrote. Adapt as needed, like change commas to apostrophes if you want.

[Stephen23]

CommaFormat('-123456') % ouch

ans = '-,123,456'

[Image Analyst]

@Stephen23 thanks for pointing that out. I've corrected it to properly handle cases where the input is negative or a string or a character array instead of a number (double, etc.). New code is attached.

It's definitely more lines than your one liner regexp though. However I'm not as adept with regexp as you -- I never would have figured out that cryptic sequence of regexp characters as you did.