csvread doesn't work, I just get errors.

I'm trying to import data from a data logger in csv format, this site says

"M = csvread('FILENAME',R,C)" is how you do that but I tried "m=csvread('LOGGER02.CSV',C5); disp(m);", but all I get is

Undefined function or variable 'C5'.

Error in untitled (line 1) m=csvread('LOGGER02.CSV',C5);

whats the real way to read csv files, assuming Matlab can read/import them? I'm using R2017b

Answer 1

[Aarti Dwivedi]

R = 1
C5 = 1
m=csvread('LOGGER02.CSV',R, C5); disp(m);

The error is pretty straightforward that your variable definitions are missing. Just saying C5 doesn't mean that the cell address is C5. C5 is the name of the variable in which you will store the column offset.

[Hank Gunderson]

I'm just going off the help section that says

"M = csvread(filename,R1,C1) reads data from the file starting at row offset R1 and column offset C1."

I don't see anything about how to define rows/columns. is there a way to read a specific column? the help section makes it sound like putting in C1 makes it read column 1.

[Walter Roberson]

csvread(filename,R1,C1) means that you want to start reading with the upper left corner skipping rows by row count given in the R1 position, and skipping columns by column count given in the C1 position. For example to have the upper left corner be row 7 column 3, then you would use csvread(filename, 6, 2) -- skipping 6 rows and 2 columns.

This does not use spreadsheet notation. The variable C1 does not mean the same as excel spreadsheet location 'C1'

If you want to specify the location 'C5' in excel spreadsheet notation, then use

csvread(filename, [], [], 'C5')

[Aarti Dwivedi]

You are supposed to put the row number in R1, and column number in C1, R1 and C1 don't mean "row 1" or "column 1" or cell number "R1" or cell number "C1"

[Hank Gunderson]

so writing:

m=csvread('LOGGER02.CSV',5);

would make it read all of column 5 right?

[Walter Roberson]

No, csvread('LOGGER02.CSV',5) would be equivalent to csvread('LOGGER02.CSV',5,0) which would skip rows 1 through 5 and read everything below that.

If you want to read only column 5, you can use the range parameter

csvread('LOGGER02.CSV', [], [], [0 4 0 inf])

Note: what happens in these cases is that csvread() uses textscan to skip the number of rows needed by using HeaderLines, and ignores the appropriate number of leading columns in each row (finding their endpoint by ignoring their content). But everything after that is read in, and the unwanted parts are thrown away afterwards. You cannot use csvread() range parameters to avoid reading the parts of the file after a given row or column, such as for performance reasons, or in an attempt to avoid reading trailing text strings.

[Hank Gunderson]

it says "Error using csvread (line 35) File not found.

Error in untitled (line 1) csvread('LOGGER02.CSV', [], [], [0 4 0 inf]) " do I have to find the file for it?

[Walter Roberson]

If you are using MS Windows then

!cd C:\; dir /s LOGGER02.CSV

to locate the file.

[Hank Gunderson]

What’s the command for Mac?

[Walter Roberson]

!find / -depth -type f -name LOGGER02.CSV -print

To restrict the search to your own files then

    !find ~ -depth -type f -name LOGGER02.CSV -print

Reminder: Mac filesystems are permitted to be case sensitive. Depending on how you configured your individual filesystem, the above might or might not match a file named (for example) Logger02.csv

I routinely configure my Mac filesystems to be case sensitive, except for one that I reserve for use with buggy programs that have mismatches in how they talk about particular files.

[Hank Gunderson]

ok I dragged the csv file into Matlab and it not crying about that anymore but now its "error" message is:

Operands to the || and && operators must be convertible to logical scalar values.

Error in dlmread (line 107)
    if r > range(3) || c > range(4), result= []; return, end

Error in csvread (line 50)
    m=dlmread(filename, ',', r, c, rng);

Error in untitled (line 1)
m=csvread('LOGGER02.CSV', [], [], [0 4 0 inf]);

non of those && in my code

m=csvread('LOGGER02.CSV', [], [], [0 4 0 inf]);
disp(m);

im not sure Matlab is even able to do this, at least mine isn't.

[Hank Gunderson]

ps I tried putting in number like this:

m=csvread('LOGGER02.CSV', 1, 2, [0 4 0 inf]); disp(m);

and nothing happens, it just says untitled followed by blank space

[Walter Roberson]

m = csvread('LOGGER02.CSV', 0, 4, [0 4 0 inf]);

[Hank Gunderson]

I used

m=csvread('LOGGER02.CSV', 1, 4, [0 4 0 inf]); disp(m);

because row one is characters, but it won't read anything, it returns nothing and the variable m is an empty 0X0 double. I just checked it and the fifth column has numbers and it imports as a 4x6 table so it should get something

[Walter Roberson]

 m = csvread('LOGGER02.CSV', 1, 4, [1 4 inf 4]); disp(m);

[Hank Gunderson]

now it says:

"Error using dlmread (line 165) Internal size mismatch

Error in csvread (line 50) m=dlmread(filename, ',', r, c, rng);

Error in untitled (line 1) m = csvread('LOGGER02.CSV', 1, 4, [1 4 inf 4]);"

[Walter Roberson]

You know, it would be a lot easier to use readtable() especially with detectImportOptions .

[Hank Gunderson]

can you use that and get just the data you want? for example if I just want the temperature column and humidity column can I make it read just those or once read, can I take out just those? I'm trying to get just those two. I'm using csv files written by a data logger

[Walter Roberson]

The simplest way is

   t = readtable('LOGGER02.CSV', 'HeaderLines', 1, 'ReadVariableName', false);
   temperature = t{:,2}; %or as appropriate
   humidity = t{:,4};    %or as appropriate

If you have R2016b or later, you can use detectImportOptions and set the SelectedVariables property to cause it to throw away everything else when you use readtable()

For older versions of MATLAB, before R2013b, then probably it would be best to use fopen()/textscan()/fclose() . This requires knowing the format of the columns you are skipping.

[Hank Gunderson]

That works great thanks!

Answer 2

[Image Analyst]

If "C5" is the column/row you see when you open it in Excel, then row = 5 and column = 3, and csvread() expects integers for offsets, not actual row or column numbers (i.e. 4 and 2):

Starting row offset, specified as a nonnegative integer. The first row has an offset of 0.
Starting column offset, specified as a nonnegative integer. The first column has an offset of 0.

so try

m=csvread('LOGGER02.CSV', 4, 2);