Re: How to extract variable from table keeping the floating point?

Rita, {} on a table returns one array. It's likely that your table
contains both integer and double variables. {} subscripting has to
concatenate all the variables, and in MATLAB, integer types are dominant.

The solution would be to pull out the doubles and the integers
separately, convert the integers to double, and concatenate. Hope this
helps.


On 7/1/2015 11:22 AM, Rita Angélico wrote:
> Hello,
>
> I am quite new to the table data type and I have a table of values of
> feautures extracted from an image, I need to convert it to a matrix
> before performing classification on it. I extract the data from the
> table like this:
> observations=msr{1:height(msr),{'DNA_Area','DNA_Eccentricity','DNA_Perimeter','DNA_MeanIntensity','DNA_MaxIntensity','DNA_StandardDeviation','DNA_Mass','DNA_P2A'}}
> ;
>
> The problem is that I rounds the values to the closest integer, and most
> of the variables are a value between [0,1] which means I need it to be
> in floating point format and not a rounded value. How can I overcome
> this problem?
>
> Thanks in advance,