Re: dataframe construction question (part 2)

[shlagbaum]

The new version of dataframe 10.1 do fixes this issues. However, I noticed
some other inconsistencies in dataframe design and I'd appreciate it a lot
if somebody could explain them to me. In particular, are there any
restrictions on column names? A simple example follows.

This works fine:
octave:23> y=dataframe([1,2,3;4,5,6],'colnames',{'c1','c2','c3'});
octave:24> y.c1
ans =

   1
   4

octave:25> y.c2
ans =

   2
   5

octave:26> y.c3
ans =

   3
   6

Now let's do the same with different column names.

octave:27> y=dataframe([1,2,3;4,5,6],'colnames',{'a','b','c'});
octave:28> y.a
error: The output format qualifier 'as' must be followed by a type
error: called from:
error:   /home/slava/octave/dataframe-1.0.1/@dataframe/subsref.m at line 45,
column 9

What? What's wrong with a column name being 'a'? And if for some reason this
name is unacceptable for dataframe constructor why it silently accepts it?

Column 'b' seems to be fine:
octave:28> y.b
ans =

   2
   5

However, the column 'c' looks really bizarre:
octave:29> y.c
ans =

   1   2   3
   4   5   6

Thanks.



--
View this message in context: 
http://octave.1599824.n4.nabble.com/dataframe-construction-question-tp4666408p4666623.html
Sent from the Octave - General mailing list archive at Nabble.com.