Thread Subject: xlsread modification

If excel file, say temp.xls, has data like this,

'A' 10
'B' 20

and if I use

[dNum, dTxt] = xlsread('temp.xls');

It gives me dNum = [10;20]

I want MATLAB to give me [nan 10; nan 20]. I tried to use a
function which overwrites MATLAB but it gets complicated.

Anybody knows easier way to solve this problem.

I checked xlsread function apparently it has data in the
format that I need but it later chops NaNs from the first
columns. I don't want to changed default xlsread.

Any solutions will be appreciated.

Chaos

Chaos <achordedrose@gmail.com> wrote:
> If excel file, say temp.xls, has data like this,
>
> 'A' 10
> 'B' 20
>
> and if I use
>
> [dNum, dTxt] = xlsread('temp.xls');
>
> It gives me dNum = [10;20]
>
> I want MATLAB to give me [nan 10; nan 20]. I tried to use a
> function which overwrites MATLAB but it gets complicated.
>
> Anybody knows easier way to solve this problem.
>
> I checked xlsread function apparently it has data in the
> format that I need but it later chops NaNs from the first
> columns. I don't want to changed default xlsread.
>

Later versions of xlsread (MATLAB 7 onwards (yuck)) running under Windows
(yuck) where a copy of Excel (yuck) is installed have three return
arguments available. The first two are as you used them, the third is a
cell array of all the data, both numeric and text. You can use this third
return argument as the basis of your own code. I do something a bit like
this:

[data,s1,cc] = xlsread(ffn,sn);
[cm,cn] = size(cc); % Size of cell array of all data
ncc = cellfun(@isnumeric,cc); % Find the numeric data in cc
nonnancc = ncc; % Numeric and not nan
nonnancc(ncc) = ~cellfun(@isnan,cc(ncc)); % Fill it in



--
Dr Tristram J. Scott
Energy Consultant

In article <lgL6k.5591$Nn.1451@newsfe09.ams2>,
tristram.scott@ntlworld.com says...
> Chaos <achordedrose@gmail.com> wrote:
> > If excel file, say temp.xls, has data like this,
> >
> > 'A' 10
> > 'B' 20
> >
> > and if I use
> >
> > [dNum, dTxt] = xlsread('temp.xls');
> >
> > It gives me dNum = [10;20]
> >
> > I want MATLAB to give me [nan 10; nan 20]. I tried to use a
> > function which overwrites MATLAB but it gets complicated.
> >
> > Anybody knows easier way to solve this problem.
> >
> > I checked xlsread function apparently it has data in the
> > format that I need but it later chops NaNs from the first
> > columns. I don't want to changed default xlsread.
> >
>
> Later versions of xlsread (MATLAB 7 onwards (yuck)) running under Windows
> (yuck) where a copy of Excel (yuck) is installed have three return
> arguments available. The first two are as you used them, the third is a
> cell array of all the data, both numeric and text. You can use this third
> return argument as the basis of your own code. I do something a bit like
> this:
>
> [data,s1,cc] = xlsread(ffn,sn);
> [cm,cn] = size(cc); % Size of cell array of all data
> ncc = cellfun(@isnumeric,cc); % Find the numeric data in cc
> nonnancc = ncc; % Numeric and not nan
> nonnancc(ncc) = ~cellfun(@isnan,cc(ncc)); % Fill it in
>
>
>
>

Without trying to ignite flames, any chance you'd share with us why you
say "... (MATLAB 7 onwards (yuck))..."?

--
Loren
http://blogs.mathworks.com/loren/

Loren Shure <loren@mathworks.com> wrote in message
>
> Without trying to ignite flames, any chance you'd share
with us why you
> say "... (MATLAB 7 onwards (yuck))..."?


I guess you don't have issues with Microsoft and Exel yucks.
Those yucks are universal. Although I personally don't have
any issues with Matlab 7 & onwards. I know many people who
don't like some features in new Matlab.

1. It has lot of GUI stuff which slows the computers.
2. It doesn't have webserver.

V

In article <g3ggr6$si9$1@fred.mathworks.com>,
Chaos <achordedrose@gmail.com> wrote:

>I know many people who
>don't like some features in new Matlab.

>1. It has lot of GUI stuff which slows the computers.
>2. It doesn't have webserver.

This samosa I brought for my lunch doesn't have a webserver either,
but I intend to enjoy eating it anyhow.

Perhaps you could expand upon your webserver comment, indicating
more clearly what functionality is missing and why that is a significant
problem?
--
  "Nothing recedes like success." -- Walter Winchell