Well I don't know procedural SQL so I'm using matlab to find and index things.
The problem:
I had a big table in a database and you want to fill in a certain column with an index that depends on other relationships in that table. (Basically, filling in parent-child relationships but each row has several identifiers: name, unique ID, other unique ID.)
The first loop demonstrates how to use "find" with strings: find the index in the array "name" where the cell value (string) equals input_name(i).
for i = 1:length(input_name)
igsn_output(i) = igsn(find(ismember(name, input_name(i))==1))
end
igsn_output = transpose(igsn_output);
This second one is just weird, it was giving string/cell/ etc. errors but this worked (input_parent was a number but in a cell array).
for i = 1:length(input_name)
parent_id_output{i} = cell2mat(sample_id(find(ismember(igsn, input_parent(i))==1)))
end
parent_id_output = transpose(parent_id_output);
I pasted it all into excel and then used "concatenate" to form the SQL statements.
This all could've been done in the SQL environment if I knew procedural postgreSQL.
They shouldn't let scientists code.
