Join or Merge Data

Hi, I have two large cells that I need to effectively "JOIN" together via two fields (one text and one numeric). Say one cell has 3 fields ID1(text),Date1(numeric),DataElement1(numeric). Say the second cell has 3 fields ID2, Date2, DataElement2. I need to have a third cell with a left outer join from cell1->cell2 on Id1->Id2 and Date1->Date2. This effectively will return me cell number one with an additional column (DataElement2).

What is the most effective way to go about doing this? Would it be turning the cell's into data-sets and performing the join function? I can easily write a loop to go one security at a time but because of the nature of text search functions (strcmp) this is incredibly slow when I have a million records or so to perform the search on.

Any suggestions would be much appreciated - thanks! Brian

[Brian]

I tested the dataset JOIN function with test data that was roughly 100,000 rows long. Matlab jumped to using 40GB of ram (of the 48 on my machine) and my entire system has crawled to a hault. In other words, unless someone can see an error in my join code, I might need to explore something different.

Where -

 ds1 = dataset(Date1,ID1,DataElement1)
 ds2 = dataset(Date2,ID2,DataElement2)
 ds3 = join(ds1,ds2,'type','leftouter','LeftKeys',{'Date1' 'ID1'},'RightKeys',{'Date2' 'ID2'},'mergekeys',true);

Answer 1

[per isakson]

These FEX contributions might help

• Relational Join, by Stefan Schroedl
• Merge dataset arrays, by Dimitri Shvorob