[Octave-bug-tracker] [bug #54029] "Name" is read as Na by dlmread (and by xlsread) in csv file

[Rik]

Follow-up Comment #4, bug #54029 (project octave):

Thanks for testing.  Matlab is very, very particular about only reading
numeric files with dlmread.  That seems a bit too restrictive.  In Octave,
skipping headder rows of text is straightforward and this is a common
occurrence as many data files have a few lines of introductory text.

So I conclude that it would be nice if Octave were to recognize that "Name" is
not Na and substitute the specified EMPTYVAL, but it is also a low priority. 
Matlab doesn't even allow for reading a file like nan.csv, and Octave can be
made to read it by the trivial workaround of skipping the first text row.


octave:2> dlmread ('nan.csv', ',', 1, 0)
ans =

   1   2   3   4
   5   6   7   8




    _______________________________________________________

Reply to this item at:

  <http://savannah.gnu.org/bugs/?54029>

_______________________________________________
  Message sent via Savannah
  https://savannah.gnu.org/