How do you round up or down to a decimal

I want to round UP to a specific decimal location (tenths in my current need).

I am getting errors using round():

a = 6.234;
b = round( a, 1);

gives 6.2. It works, but is not UP. It rounded DOWN. So I add TieBreaker:

b = round( a, 1, TieBreaker="plusinf");

gives

Error using round
Too many input arguments.

I missed something

b = round( a, TieBreaker="plusinf");

gives

Error using round
Third input must be either 'decimals' or 'significant'.

I missed something

Any comments, corrections, alternate methods are appreciated.

Answer 1

[Image Analyst]

To round up use ceil

To round down use floor

[Ted H]

I don't see in ceil where you can handle the decimal. I see the time component only.

[Voss]

You can do this kind of thing:

a = 6.234;

% round UP to the tenths:
b = ceil(a*10)/10

b = 6.3000

a = 6.237;

% round DOWN to the hundredths:
b = floor(a*100)/100

b = 6.2300

Answer 2

[John D'Errico]

You are trying to use capabilities of round that are not present in your (older) MATLAB release.

For that code to work, you need to upgrade to a current release. But a simple call to round should still work for you.

b = round(6.234,1)

b = 6.2000

c = round(6.253,1)

c = 6.3000

Just that the option you are trying to use is a more recent capability.

[Walter Roberson]

R2022a or later it looks like. I had not noticed that they extended the capabilities!

[John D'Errico]

I am constantly being surprised, since I too often forget to read the release notes for every release.

[Ted H]

Rereading the matlab documentation, tiebreaker is only for the exact midpoint. So there is no round up or round down. This does not solve my problem. @Image Analyst solution does work, however a minor complaint is that it reduces readability (IMO).

Unrelated to the technique, I would have thought rounding up or down to a specific decimal place was a need, but I suppose not, or it was resolved by users dividing and multiplying. Matlab was first commercialized in the 80s, and not until 2014 was this need made a function, while this is standard in many other programs.

@John D'Errico your solution is just rounding. not rounding always up or always down to a specific decimal place. I might not have made this as clear as I should have. I edited the original post.