Precision in writetable()

Is there a way to use writetable() to export using only 2 decimals? To be able to directly export it to an excel file.

Answer 1

[dpb]

Sadly, no. A OutputFormat formatting string or NumericPrecision option is not available.

I presume since a table can contain any data type including cells too many possibilities are opened up that TMW hasn't wanted to venture down that road.

Does seem like worthy of some design consideration for the normal cases, though, granted...

For text files, the numeric format is long g.

[madhan ravi]

Ah :(, any suggestions dpb? Perhaps dlmwrite() to csv and readtable() and then writetable()?

[dpb]

What's the end objective file target and the data?

Perhaps rounding the data in the table first?

[Walter Roberson]

varfun() with appropriately selected input variables, with @(v) round(v,2) as the function to be applied.

[madhan ravi]

Thank you dpb and sir Walter , rounding would be appropriate for my needs.

[Willingo]

You can do this via

t = table( 1.12345, "test", 123.12345, 12.12, "12test")

t = 1×5 table
     Var1      Var2      Var3     Var4       Var5  
    ______    ______    ______    _____    ________

    1.1235    "test"    123.12    12.12    "12test"

t{:,vartype('numeric') } = round( t{:,vartype('numeric')}, 3,'significant')

t = 1×5 table
    Var1     Var2     Var3    Var4      Var5  
    ____    ______    ____    ____    ________

    1.12    "test"    123     12.1    "12test"

[Tao Wang]

For me , Willingo's answer is still not working.

I tried and it did show numbers with 2 decimals in commond . but if I writetable() to excel, the numbers format is still long.

Also , I tried num2str first ,

 num2str(1.345,'%.2')

and u will not only get the data with 2 decimals ,but also a green flag on the top left of the box.

[dpb]

Because having rounded the numbers, they still are doubles and will have machine-precision rounding that Excel will try to preserve.

All you can do is format them inside Excel; same way as MATLAB, Excel keeps everything as a double internally; it only changes how they're displayed.

The second route writes the numeric value as text in the cell which is the source of the warning highlight.

There are several user-contributed utilities to allow one to set Excel table properties on the FEX. I believe it was @Image Analyst who wrote and posted a pretty nice starter set of Excel_utils that you may search for here on Answers -- I recall making a couple extensions and adding/posting another feature or two. That's been with the year...just at the moment I've got things in a state it isn't convenient to try to go find, but will try to get back...

Answer 2

[Johannes Kalliauer]

You could use dlmwrite: https://www.mathworks.com/help/matlab/ref/dlmwrite.html

dlmwrite('yourfile.txt',t{:,:},'\t','precision','%10.2f')

[dpb]

Doesn't get OP directly to Excel as per request, though...but correct that it does allow the formatting string.

Answer 3

[Christine]

Another workaround might be executing this line for the variables inside the table.

variable = round(variable*100)/100;

[Adam Danz]

Use Y = round(X,N) to specify the number of decimal places.

Instead of

round(pi*100)/100

ans = 3.1400

use

round(pi,2)

ans = 3.1400

[dpb]

Another I've seen has been

str2double(sprintf('%.2f',pi))

ans = 3.1400