read csv data with numbers, characters and text

Greetings

Does anyone have any idea on how to import numerical data from a csv file of the following format:

STATION DATE LATITUDE LONGITUDE ELEVATION NAME PRCP PRCP_ATTRIBUTES TMAX TMAX_ATTRIBUTES TMIN TMIN_ATTRIBUTES

BC000068026 01/02/32 -18.367 21.85 1000 SHAKAWE, BC 0 ,,Q

I'm attaching a sample file

Matlab2016b

Thank you very much in advance

[Walter Roberson]

Try readtable

[Simon Lind]

thank you. I forgot to mention that I'm tryin on M2016b, which doesn'nt have readtable

[Cris LaPierre]

I suggest posting at least a portion of your file so that we can try out different approaches. Use the paperclip icon.

[Walter Roberson]

readtable was added in R2013b. In the R2016b time-frame you would very likely want to use detectImportOptions() first.

[Simon Lind]

detectImportOptions is a very good tip

[Image Analyst]

If the format is too custom and varies from file to file, you may have to write your own custom reader where you read a line at a time and parse it appropriately.

Answer 1

[Cris LaPierre]

Have you tried readtable? What didn't work about it? I get the results I would expect. It even successfully ignored the delimiters inside quotes.

The file has 12 delimited fields, the names of which are given in the first line of your file.

The "SHAKAWE, BC" AND ",,Q" are each considered single variables (NAME and PRCP ATTRIBUTES). When I use readtable in R2016b with no options, I get this:

Notice the warning about dates. I suspect it should be dd/MM/yy. Use the options to fix that.

opts = detectImportOptions('sample.csv');
opts = setvaropts(opts,'DATE','InputFormat','dd/MM/yy');
data = readtable('sample.csv',opts)

[Cris LaPierre]

There are a lot of empty lines. It doesn't appear the 'EmptyLineRule' does anything in R2016b, but you can delete the emptry rows using this code.

idx = all(ismissing(data),2);
data(idx,:)=[];

[Simon Lind]

I followed your advices. This was what I finally did and it works

Thank you very much

opts.SelectedVariableNames = {'DATE','LATITUDE','LONGITUDE','PRCP'};
T = readtable([datapath myfile],opts);
t = table2array(T(:,1));
lat = table2array(T(:,2));
lon = table2array(T(:,3));
p = table2array(T(:,4));
lon = str2num(cell2mat(lon));
lat = str2num(cell2mat(lat));
t = datenum(t);
p = str2num(cell2mat(p));

[Cris LaPierre]

Why all the postprocessing? See how to access data in tables.