Read numbers and characters

Hi!

I am trying to read an output from a software that gives me numbers and semicolons, commas, minus and points. The output can be either an excel file ou txt file. If I choose the excel file and open it in matlab, the output is like this:

 1
9
.
3
1
2
5
COMMA
2
5
.
1
2
5
0
COMMA
-
1
3
.
7

But I want to matlab recognize if it is a number or a character. In reality, what I want to do is to have "19.3125, 25.1250, -13.7,..." and for that I need matlab to recognize what is a number and not. Can you help me?

[Jan]

What does "open the file in Matlab" exactly mean? By fopen, in the editor, imported by fileread or importdata? Does "COMMA" mean a comma or literally the string 'COMMA'? Are there linebreaks after each character?

[Stephen23]

@Patrícia Falcão: please upload a sample text file by clicking the paperclip button.

[Patrícia Falcão]

It means I imported to matlab, there is an image. COMMA means ",". I want to create a for cycle and then, whenever "C" appears, I want to transform it into a new column.

[Stephen23]

'COMMA means ","'

Sure, we know what "comma" means. But which of these is actually in the text file?:

1. ,
2. COMMA

This would be much simpler if you just uploaded a sample text file by clicking on the paperclip button.

[Patrícia Falcão]

You're right. It is uploaded the txt file. But I don't want the first numbers, just the next to the first comma. What I have already explained to Stephen Cobeldick.

Answer 1

[Stephen23]

Using a simple text file (this will be much easier than processing an Excel file):

>> str = fileread('temp0.txt');
>> str = regexprep(str,'\s+','');
>> vec = sscanf(str,'%fCOMMA',[1,Inf])
vec =
   19.312   25.125  -13.700

The test file is attached. If you had actually uploaded your file then I could have tested my code on that: no sample data means I have to invent my own data, which may or may not be the same as what you have.

[Patrícia Falcão]

That works perfectly! The only problem is that the output in txt file is another one:

%
Time [s],Value,Parity Error,Framing Error
0.050669000000000,1,,
0.050754000000000,9,,
0.050840000000000,.,,
0.050925000000000,3,,
0.051010000000000,1,,
0.051095000000000,2,,
0.051180000000000,5,,
0.051265000000000,COMMA,,
0.051350000000000,2,,
0.051435000000000,5,,
0.051520000000000,.,,
0.051605000000000,1,,
0.051690000000000,2,,
0.051775000000000,5,,
0.051860000000000,0,,
0.051945000000000,COMMA,,
0.052030000000000,1,,
0.052115000000000,3,,
0.052200000000000,.,,
0.052285000000000,7,,
0.052370000000000,5,,
0.052455000000000,0,,
0.052540000000000,0,,
0.052625000000000,COMMA,,
0.052710000000000,3,,
0.052795000000000,9,,
0.052880000000000,.,,
0.052965000000000,8,,
0.053050000000000,7,,
0.053135000000000,5,,
0.053220000000000,0,,
0.053305000000000,COMMA,,
0.053390000000000,0,,
0.053475000000000,.,,
0.053560000000000,8,,
0.053645000000000,1,,
0.053730000000000,2,,
0.053815000000000,5,,
0.053900000000000,COMMA,,
0.053985000000000,-,,
0.054070000000000,1,,
0.054155000000000,.,,
0.054240000000000,5,,
0.054325000000000,6,,
0.054410000000000,2,,
0.054495000000000,5,,
0.054580000000000,;,,
0.054665000000000,\r,,
0.054750000000000,\n,,
0.156414000000000,1,,
0.156499000000000,9,,
0.156584000000000,.,,
0.156669000000000,3,,
0.156754000000000,1,,
0.156839000000000,2,,
0.156924000000000,5,,
0.157009000000000,COMMA,,

I don't want the first number (time), I just want the next char/number next to the first comma. In this case, my final goal is to have something like that: 19.3125, 25.1250, 13.7500, 39.8750, 0.8125, -1.5625; (then a new line) 19.3125, ...

[Stephen23]

You can read the second column of that text file like this:

>> fmt = '%*s%s%*[^\n]'; % keep only second column.
>> opt =  {'Delimiter',',','HeaderLines',1};
>> [fid,msg] = fopen('channel3txt.txt','rt');
>> assert(fid>=3,msg)
>> C = textscan(fid,fmt,opt{:}); % read file data into MATLAB
>> fclose(fid);
>> C = strrep(C{1},'COMMA',','); % replace COMMA -> ,
>> str = sprintf([C{:}]) % convert to char vector, and \r\n -> to newline characters.
str = 19.3125,25.1250,13.7500,39.8750,0.8125,-1.5625;
19.3125,25.1250,13.7500,39.8750,0.8125,-1.5625;
19.3125,25.1250,13.7500,39.8750,0.8125,-1.5625;
19.3125,25.1250,13.7500,39.8750,0.8125,-1.5625;
19.3125,25.1250,13.7500,39.8750,0.8125,-1.5625;
... lots of rows here
66.3125,48.6250,165.8750,138.6250,27.8125,175.7500;
57.8125,59.2500,148.1875,1.7500,18.8750,113.4375;
50.0625,62.0000,96.2500,347.9375,2.1875,60.0625;
48.1250,59.5000,92.5000,345.5625,0.6250,63.9375;
59.7500,61.1250,144.8125,2.1875,18.6250,119.5625;
69.3750,46.9375,168.6250,139.4375,21.7500,171.3750;
62.2500,51.3750,164.8750,13

Note that your sample file does not include data for the complete last row, it appears to be truncated.

At this point str is a char vector with numbers separated by commas and semi-colons with newlines. You can efficiently convert that char vector into a numeric vector using:

>> vec = sscanf(str,'%f,%f,%f,%f,%f,%f;')
vec =
    19.31250
    25.12500
    13.75000
    39.87500
     0.81250
    -1.56250
    19.31250
    25.12500
... lots of rows here:
    46.93750
   168.62500
   139.43750
    21.75000
   171.37500
    62.25000
    51.37500
   164.87500
    13.00000

At that point you need to decide what to do with the incomplete last row. One option is to ignore it, allowing you to reshape the data into this numeric matrix:

>> vec = vec(1:6*fix(numel(vec)/6)); % remove incomplete last row
>> mat = reshape(vec,6,[]).'
mat =
    19.31250    25.12500    13.75000    39.87500     0.81250    -1.56250
    19.31250    25.12500    13.75000    39.87500     0.81250    -1.56250
    19.31250    25.12500    13.75000    39.87500     0.81250    -1.56250
    19.31250    25.12500    13.75000    39.87500     0.81250    -1.56250
    19.31250    25.12500    13.75000    39.87500     0.81250    -1.56250
    19.31250    25.12500    13.75000    39.87500     0.81250    -1.56250
    19.31250    25.12500    13.75000    39.87500     0.81250    -1.56250
    19.31250    25.12500    13.75000    39.87500     0.81250    -1.56250
 ... lots of rows here
    50.06250    62.00000    96.25000   347.93750     2.18750    60.06250
    48.12500    59.50000    92.50000   345.56250     0.62500    63.93750
    59.75000    61.12500   144.81250     2.18750    18.62500   119.56250
    69.37500    46.93750   168.62500   139.43750    21.75000   171.37500

[Patrícia Falcão]

Thank you very much Stephen Cobeldick! You helped me a lot!

[Stephen23]

@Patrícia Falcão: I hope that it helps! Remember to accept the answer that helped you most: this give reputation points to the volunteers who help you on this forum.