Compare two tables with two column combinations

I have two tables, say:
t1 = table(categorical({'a';'b';'c';'c';'a';'b'}),categorical({'d';'e';'d';'f';'e';'d'}),[1;2;3;4;5;6],datetime({'2017-1-1 12:00 AM','2017-1-3 10:20 AM','2017-1-3 1:36 PM','2017-1-4 9:45 AM','2017-1-6 12:00 AM','2017-1-7 12:12 PM'})')
t2 = table(categorical({'a';'a';'c';'b';'c'}),categorical({'f';'d';'e';'f';'d'}),[5;4;3;2;1],datetime({'2017-1-1 12:00 AM','2017-1-3 6:20 AM','2017-1-3 1:36 PM','2017-1-4 9:45 AM','2017-1-6 12:10 AM'})')
t1 =
6×4 table
Var1 Var2 Var3 Var4
____ ____ ____ ____________________
a d 1 01-Jan-2017 00:00:00
b e 2 03-Jan-2017 10:20:00
c d 3 03-Jan-2017 13:36:00
c f 4 04-Jan-2017 09:45:00
a e 5 06-Jan-2017 00:00:00
b d 6 07-Jan-2017 12:12:00
t2 =
5×4 table
Var1 Var2 Var3 Var4
____ ____ ____ ____________________
a f 5 01-Jan-2017 00:00:00
a d 4 03-Jan-2017 06:20:00
c e 3 03-Jan-2017 13:36:00
b f 2 04-Jan-2017 09:45:00
c d 1 06-Jan-2017 00:10:00
Now I want the result or the rownumbers of every combination of the first two columns in t2 that are not in t1
How can I do this?
*Edited question with a little more complicated table

 Réponse acceptée

tables have all the same "set membership" functions as numeric arrays:
>> t1 = table({'a';'b';'c';'c';'a';'b'},{'d';'e';'d';'f';'e';'d'});
>> t2 = table({'a';'a';'c';'b';'c'},{'f';'d';'e';'f';'d'});
>> [t3,i1] = setdiff(t1,t2)
t3 =
4×2 table
Var1 Var2
____ ____
'a' 'e'
'b' 'd'
'b' 'e'
'c' 'f'
i1 =
5
6
2
4

3 commentaires

Aletta Wilbrink
Aletta Wilbrink le 7 Fév 2018
Modifié(e) : Aletta Wilbrink le 7 Fév 2018
Thank you.
But what if I have a third column (with numbers or dates) and I don't want that one in the comparison?
Well, then you use
[~, i1] = setdiff(t1(:, [1 2]), t2(:, [1 2]));
t3 = t1(i1, :)
Hi Peter,
Your answer is very elegant, thank you. Do you have any thoughts please on also checking for differences in value vs. not-a-value status -- i.e. NaN, NaT ?
Would you recommend working along the lines of e.g. varfun(@isnan..)) ?
Thanks,
Brad

Connectez-vous pour commenter.

Plus de réponses (1)

If I understand what you want to do, this will work:
Lv = ismember(t2, t1, 'rows');
DesiredRows = find(~Lv)
DesiredRows =
1
3
4
If not, reversing the arguments to ismember (link) will work.

Catégories

Tags

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by