Use Import's automatic table detection programatically

When opening a *.csv with header in the "Import" window, the data table is automatically selected. Is there a way this can be done programatically? The methods described elsewhere (importdata, textread, csvread etc) are less generic and require knowledge of either the number of headerlines or header/table contents.

[Walter Roberson]

Could you post an example of the first few lines of a couple of representative files you need to handle?

Answer 1

[Jeremy Hughes]

It sounds like the R2016b function, http://www.mathworks.com/help/matlab/ref/detectimportoptions.html

Answer 2

[Sean de Wolski]

Have you tried readtable?

[C Hinderink]

Yes, it returns: "... Found 2 variable names but 1 data columns. You may need to specify a different format string, delimiter, or number of header lines." So not the auto-table detect the GUI provides...

[Sean de Wolski]

Also, give http://www.mathworks.com/help/matlab/ref/delimitedtextimportoptions-object.html in R2016b.

Answer 3

[Star Strider]

I don’t know what you included with ‘etc.’ in the file import functions you’ve used. Have you tried xlsread? It might be able to read your .csv file and do everything you want.

[C Hinderink]

xlsread does about the same. Alas the two downsides are that it returns one matrix and only numeric values.

[Star Strider]

You can get three outputs from xlsread, the first being the numeric values, the second all strings (column headers, dates, string variables), and the third everything in the file as a cell array. See Request Numeric, Text, and Raw Data in the xlsread documentation for details.