Matlab read .csv file together with headlines and operations on matrix elements?

Hi to everyone,

I need help regarding reading csv file where Ill be able to access and read headlines:

this is the csv file and I need first row, headline too.

freq Unit1 Unit2 Unit3 Unit4 Unit5 Unit6 Unit7 Unit8 Unit9 Unit10

50 2.0614 4.5103 3.918 3.4831 1.56 4.884 1.621 3.9072 1.6264 3.9601

150 1.811 3.907 2.4978 3.3853 2.42 3.397 3.7612 2.1432 0.67351 1.566

250 1.4037 0.67744 1.4482 0.74906 0.64507 2.575 0.6273 0.95441 0.535 2.2526

My code is:

filename1 = 'DBF.csv';

DB = readtable(filename1,'ReadVariableNames', false);

But when I am trying to read headline, I am getting this:

Unit1

_____

1.811

Now I am confused as I cant work with numbers nor with strings? How to get only values and strings from headline?

Second question is regarding IF statement where I am taking elements of matrix to test:

nrows=3; ncols=11;

for c=2:ncols

det=0;

for r=1:nrows

low=LM*DB(r,c); high=HM*DB(r:c); Here I am getting the error most likely as DB(r,c) is as I show.

if low < NU(r:2) && NU(r:2) < high

det=det+1;

else det=0;

end

if det==3

fprintf('Newly connected unit is %d\n', DB(c:1));

end

end

end

if det==0

fprintf('Newly connected unit cannot be rcognised! %d\n');

end

Hopefully someone can help in this.

Many thanks in advance.

Answer 1

[Star Strider]

The readtable function call may be in error if you want to import the variable names, however I don’t know from the code what you want to do.

I would do something like this —

DB = readtable(filename1);

VN = DB.Properties.VariableNames

Ths second line saves them to ‘VN’ as a cell array. You can do whatever you want to with them after that. (The current version of readtable also supports 'VariableNamingRule','preserve' to read variable names that are not vallid MATLAB variable names. A previous version of that exists as 'PreserveVariableNames',true , however I don’t remember when it was introduced. It may not be important here if the variable names are valid MATLAB variable names.)

[Mike Pierie]

Hi Star Sider and thanks for the quick answer.

I would like to import those data into Matlab in order to make some comparsion of values and as output should be: Unit match with existing one or doesnt match.

Ill try what you told me.

[Mike Pierie]

I found somewhere yours comment and recommendation to use "readcell" here is what I am getting in both cases:

readcell:

Columns 1 through 9

{'freq'} {'Unit1' } {'Unit2' } {'Unit3' } {'Unit4' } {'Unit5' } {'Unit6' } {'Unit7' } {'Unit8' }

{[ 50]} {[2.0614]} {[4.5103]} {[3.9180]} {[3.4831]} {[1.5600]} {[4.8840]} {[1.6210]} {[3.9072]}

{[ 150]} {[1.8110]} {[3.9070]} {[2.4978]} {[3.3853]} {[2.4200]} {[3.3970]} {[3.7612]} {[2.1432]}

{[ 250]} {[1.4037]} {[0.6774]} {[1.4482]} {[0.7491]} {[0.6451]} {[2.5750]} {[0.6273]} {[0.9544]}

Columns 10 through 11

{'Unit9' } {'Unit10'}

{[1.6264]} {[3.9601]}

{[0.6735]} {[1.5660]}

{[0.5350]} {[2.2526]}

and VN = DB.Properties.VariableNames;

Columns 1 through 9

{'freq'} {'Unit1'} {'Unit2'} {'Unit3'} {'Unit4'} {'Unit5'} {'Unit6'} {'Unit7'} {'Unit8'}

Columns 10 through 11

{'Unit9'} {'Unit10'}

Can I use that for the purpose of my code?

[Star Strider]

My pleasure!

I didn’t specifically mention readcell here, although I have in answers to other questions. The first row of the readcell result should produce the same as the ‘VN’ variable with readtable.

I would use readtable for this, however I’m still not certain that I understand what you want to do. If you want to extract the data from the ‘DB’ table use the table2array function, although it is relatively straightforward to address the individual variables (columns) of the table by using the variable names, as well as referencing them as cells with numerical indices.

With respect to the loop in your original post, the element references in the if block (and elsewhere as necessary) should likely be to ‘DB{r,c}’ using curly brackets instead of parentheses, since ‘DB’ is a table.

EDIT — (24 Aug 2022 at 15:48)

I didn’t see the file you posted until just now.

DB = readtable('https://www.mathworks.com/matlabcentral/answers/uploaded_files/1106800/DBF.csv')

DB = 3×11 table
    freq    Unit1      Unit2     Unit3      Unit4      Unit5     Unit6    Unit7      Unit8      Unit9     Unit10
    ____    ______    _______    ______    _______    _______    _____    ______    _______    _______    ______

     50     2.0614     4.5103     3.918     3.4831       1.56    4.884     1.621     3.9072     1.6264    3.9601
    150      1.811      3.907    2.4978     3.3853       2.42    3.397    3.7612     2.1432    0.67351     1.566
    250     1.4037    0.67744    1.4482    0.74906    0.64507    2.575    0.6273    0.95441      0.535    2.2526

VN = DB.Properties.VariableNames

VN = 1×11 cell array
    {'freq'}    {'Unit1'}    {'Unit2'}    {'Unit3'}    {'Unit4'}    {'Unit5'}    {'Unit6'}    {'Unit7'}    {'Unit8'}    {'Unit9'}    {'Unit10'}

FirstVariable = VN{1}

FirstVariable = 'freq'

LastVariable = VN{end}

LastVariable = 'Unit10'

So if you need the variable names as character arrays, just use cell addressing to return them as such.

.

[Mike Pierie]

aaaa great!!! {r,c} this is what I needed! Now it reading elements as numbers, great!

Now next point is:

if low < NU{r,2} && NU{r,2} < high

in the loop where I am getting error: Operands to the || and && operators must be convertible to logical scalar values.

for c=2:ncols

det=0;

for r=1:nrows

low=LM*C{r,c}; high=HM*C{r,c};

if low < NU{r,2} && NU{r,2} < high

det=det+1;

else det=0;

end

if det==3

fprintf('Newly connected unit is %d\n', C{1,c});

end

end

end

if det==0

fprintf('Newly connected unit cannot be rcognised! %d\n');

end

Also how I can acess now and use the names, strings Unit1,2,3.. after VN = DB.Properties.VariableNames; ?

Many many thanks to both of you!

[Star Strider]

Ths ‘short circuit’ operators only operate on logical scalars.

Try this —

DB = readtable('https://www.mathworks.com/matlabcentral/answers/uploaded_files/1106800/DBF.csv')

DB = 3×11 table
    freq    Unit1      Unit2     Unit3      Unit4      Unit5     Unit6    Unit7      Unit8      Unit9     Unit10
    ____    ______    _______    ______    _______    _______    _____    ______    _______    _______    ______

     50     2.0614     4.5103     3.918     3.4831       1.56    4.884     1.621     3.9072     1.6264    3.9601
    150      1.811      3.907    2.4978     3.3853       2.42    3.397    3.7612     2.1432    0.67351     1.566
    250     1.4037    0.67744    1.4482    0.74906    0.64507    2.575    0.6273    0.95441      0.535    2.2526

VN = DB.Properties.VariableNames

VN = 1×11 cell array
    {'freq'}    {'Unit1'}    {'Unit2'}    {'Unit3'}    {'Unit4'}    {'Unit5'}    {'Unit6'}    {'Unit7'}    {'Unit8'}    {'Unit9'}    {'Unit10'}

% FirstVariable = VN{1}
% LastVariable = VN{end}

LM =  1;                                % Create These
HM = 10;

[nrows,ncols] = size(DB)

nrows = 3

ncols = 11

 for c=2:ncols 
     det=0;
    for r=1:nrows 
        low=LM*DB{r,c}; 
        high=HM*DB{r,c};
        if low < NU{r,2} & NU{r,2} < high
        det=det+1;      
        else det=0;
        end
        if det==3
            fprintf('Newly connected unit is %d\n', DB(c:1));
        end
    end
 end

Undefined variable 'NU'.

 if det==0
     fprintf('Newly connected unit cannot be rcognised! %d\n');
 end

I have no idea what ‘NU’ is, however its absence now appears to be the only problem with the code.

.

[Mike Pierie]

Once more again, thanks to the sky!

Almost done and here is the code:

clc;

clear all;

close all;

filename1 = 'DBF.csv'; % parameters for all recorded units

filename2 = 'PNU.csv'; % parameters of new unit

LM=0.8; % Margin 1 - 20% below

HM=1.2; % Margin 2 - 20% above

% C = readcell(filename1);

% disp(C);

DB = readtable(filename1,'ReadVariableNames', true); % parameters of all recorded units

VN = DB.Properties.VariableNames;

[nrows,ncols] = size(DB);

NU = readtable(filename2,'ReadVariableNames', true); % parameters of newly connected unit

VNNU = NU.Properties.VariableNames; % is this correct?

for c=2:ncols

det=0;

for r=1:nrows

low=LM*DB{r,c};

high=HM*DB{r,c};

newunitH=NU{r,2};

if low < newunitH & newunitH < high

det=det+1;

else det=4

end

disp(det);

if det==3

% at 1st pass det is 3 but fprint doesnt print it well. disp(VN{c}) showing correct header.

% Most likely I put in

% wrong way VN{c} in fprint as is showing following:

% Newly connected unit is 85

% Newly connected unit is 110

% Newly connected unit is 105

% Newly connected unit is 116

% Newly connected unit is 49

% instead "Newly connected unit is Unit1"

disp(c);

disp(VN{c});

fprintf('Newly connected unit is %d\n', VN{c});

end

end

end

if det==0

fprintf('Newly connected unit cannot be rcognised! %d\n');

end

[Mike Pierie]

I realized that problem was in the %d which is for integers and % should be for the strings... Now working everyhing!!!

Guys I am so thankful to you, especially to Mr. Star Strider.

Step 2 is done.. There will be Step 3 too and I am working on it now.

Have an excellent time there where you are!

[Star Strider]

As always, my pleasure!

Answer 2

[Cris LaPierre]

Please share your csv file. You can attach it using the paperclip icon.

A preliminary look would suggest the code you have shared is not the same code that is used to read the csv file, as the output shows that the variable names have been read in.

As an example, I saved the file contents you shared to a csv and loaded it using your code. As you can see, it read in all the data, so I suspect there are some missing details.

filename1 = 'DBF.csv';
DB = readtable(filename1,'ReadVariableNames', false)

DB = 3×11 table
    Var1     Var2      Var3       Var4      Var5       Var6      Var7      Var8      Var9       Var10     Var11 
    ____    ______    _______    ______    _______    _______    _____    ______    _______    _______    ______

     50     2.0614     4.5103     3.918     3.4831       1.56    4.884     1.621     3.9072     1.6264    3.9601
    150      1.811      3.907    2.4978     3.3853       2.42    3.397    3.7612     2.1432    0.67351     1.566
    250     1.4037    0.67744    1.4482    0.74906    0.64507    2.575    0.6273    0.95441      0.535    2.2526

If you are unfamiliar with tables, I recommend visiting the Access data in Tables page. Or you can use readmatrix instead.

DB = readmatrix(filename1)

DB = 3×11

50.0000 2.0614 4.5103 3.9180 3.4831 1.5600 4.8840 1.6210 3.9072 1.6264 3.9601 150.0000 1.8110 3.9070 2.4978 3.3853 2.4200 3.3970 3.7612 2.1432 0.6735 1.5660 250.0000 1.4037 0.6774 1.4482 0.7491 0.6451 2.5750 0.6273 0.9544 0.5350 2.2526

[Mike Pierie]

Hi Cris and manz thanks for quick response!

I attached the file.

I found that readtable is somehow easiest but I cant get

Unit1 Unit2 Unit3 Unit4 Unit5 Unit6 Unit7 Unit8 Unit9 Unit10 as strings and the rest as numbers.

Ill check the link you provided.

[Cris LaPierre]

filename = 'https://www.mathworks.com/matlabcentral/answers/uploaded_files/1106800/DBF.csv';
DB = readtable(filename)

DB = 3×11 table
    freq    Unit1      Unit2     Unit3      Unit4      Unit5     Unit6    Unit7      Unit8      Unit9     Unit10
    ____    ______    _______    ______    _______    _______    _____    ______    _______    _______    ______

     50     2.0614     4.5103     3.918     3.4831       1.56    4.884     1.621     3.9072     1.6264    3.9601
    150      1.811      3.907    2.4978     3.3853       2.42    3.397    3.7612     2.1432    0.67351     1.566
    250     1.4037    0.67744    1.4482    0.74906    0.64507    2.575    0.6273    0.95441      0.535    2.2526

% Capturing variable names
varNms = string(DB.Properties.VariableNames)

varNms = 1×11 string array
    "freq"    "Unit1"    "Unit2"    "Unit3"    "Unit4"    "Unit5"    "Unit6"    "Unit7"    "Unit8"    "Unit9"    "Unit10"

% Extract data from one column (all values vs specific values)
DB.Unit1

ans = 3×1

2.0614 1.8110 1.4037

DB.Unit2(2:3)

ans = 2×1

3.9070 0.6774

[Mike Pierie]

Hi Cris, I ovelooked your 2nd comment and thanks again for it.

It seams with varNms = string(DB.Properties.VariableNames) I can get var names and hopefully connect with the evaluation of values. Once when Il detect correct case to print the name of appropriate Unit.

Can I use index to acces to particular elemet of DB like DB.Unit1(1,1) as input for evaluation maybe? Ill try immediatelly when Ill be home.

[Mike Pierie]

and also is it possible to automatize it in order to check entire file and not only one of the columns?

I mean Unit1 2 3 4... in DB.Unit expression.

[Cris LaPierre]

I don't know what it is you are wanting to compare, but yes, the column number is the same for the data and the variables.

For a table, use the width and height methods to get the number of columns and rows, respectively.

The page on accessing data in tables I linked to previously is helpful, as there are multiple ways to access the data. You can find the one that works best for you.

filename = 'https://www.mathworks.com/matlabcentral/answers/uploaded_files/1106800/DBF.csv';
DB = readtable(filename)

DB = 3×11 table
    freq    Unit1      Unit2     Unit3      Unit4      Unit5     Unit6    Unit7      Unit8      Unit9     Unit10
    ____    ______    _______    ______    _______    _______    _____    ______    _______    _______    ______

     50     2.0614     4.5103     3.918     3.4831       1.56    4.884     1.621     3.9072     1.6264    3.9601
    150      1.811      3.907    2.4978     3.3853       2.42    3.397    3.7612     2.1432    0.67351     1.566
    250     1.4037    0.67744    1.4482    0.74906    0.64507    2.575    0.6273    0.95441      0.535    2.2526

width(DB)

ans = 11

height(DB)

ans = 3

DB.Unit1(2) % 2nd value in Unit 1 column

ans = 1.8110

DB.Unit1(2,1)  % 2nd value in Unit 1 column

ans = 1.8110

DB{2,2}  % 2nd value in 2nd column of DB (Unit 1)

ans = 1.8110

DB.(DB.Properties.VariableNames{2})(2) % 2nd value in 2nd variable column (Unit 1)

ans = 1.8110

DB.(string(DB.Properties.VariableNames{2}))(2) % 2nd value in 2nd variable column (Unit 1)

ans = 1.8110

To check values across all variables, assuming all columns contain numeric data, you would use curly braces to extract the table data as an array (the same result you would get by using readmatrix).

db=DB{:,:} % all rows, all columns

db = 3×11

50.0000 2.0614 4.5103 3.9180 3.4831 1.5600 4.8840 1.6210 3.9072 1.6264 3.9601 150.0000 1.8110 3.9070 2.4978 3.3853 2.4200 3.3970 3.7612 2.1432 0.6735 1.5660 250.0000 1.4037 0.6774 1.4482 0.7491 0.6451 2.5750 0.6273 0.9544 0.5350 2.2526

db(2,2) % 2nd row, 2nd column

ans = 1.8110