Data queries in DataJoint comprise two distinct steps:

  1. Construct the query object to represent the required data using tables and operators.

  2. Fetch the data from query into the workspace of the host language – described in this section.

Note that entities returned by fetch methods are not guaranteed to be sorted in any particular order unless specifically requested. Furthermore, the order is not guaranteed to be the same in any two queries, and the contents of two identical queries may change between two sequential invocations unless they are wrapped in a transaction. Therefore, if you wish to fetch matching pairs of attributes, do so in one fetch call.

The examples below are based on the example schema for this part of the documentation.

DataJoint for MATLAB provides three distinct fetch methods: fetch, fetch1, and fetchn. The three methods differ by the type and number of their returned variables.

query.fetch returns the result in the form of an n ⨉ 1 struct array where n is the number of records matching the query expression.

query.fetch1 and query.fetchn split the result into separate output arguments, one for each attribute of the query.

The types of the variables returned by fetch1 and fetchn depend on the datatypes of the attributes. query.fetchn will enclose any attributes of char and blob types in cell arrays whereas query.fetch1 will unpack them.

MATLAB has two alternative forms of invoking a method on an object: using the dot notation or passing the object as the first argument. The following two notations produce an equivalent result:

result = query.fetch(query, 'attr1')
result = fetch(query, 'attr1')

However, the dot syntax only works when the query object is already assigned to a variable. The second syntax is more commonly used to avoid extra variables.

For example, the two methods below are equivalent although the second method creates an extra variable.

# Method 1
result = fetch(university.Student, '*');

# Method 2
query = university.Student;
result = query.fetch()

Fetch the primary key

Without any arguments, the fetch method retrieves the primary key values of the table in the form of a single column struct. The attribute names become the fieldnames of the struct.

keys = query.fetch;
keys = fetch(university.Student & university.StudentMajor);

Note that MATLAB allows calling functions without the parentheses ().

Fetch entire query

With a single-quoted asterisk ('*') as the input argument, the fetch command retrieves the entire result as a struct array.

data = query.fetch('*');

data = fetch(university.Student & university.StudentMajor, '*');

In some cases, the amount of data returned by fetch can be quite large. When query is a table object rather than a query expression, query.sizeOnDisk() reports the estimated size of the entire table. It can be used to assess whether running query.fetch('*') would be wise. Please note that it is only currently possible to query the size of entire tables stored directly in the database .

As separate variables

The fetch1 and fetchn methods are used to retrieve each attribute into a separate variable. DataJoint needs two different methods to tell MATLAB whether the result should be in array or scalar form; for numerical fields it does not matter (because scalars are still matrices in MATLAB) but non-uniform collections of values must be enclosed in cell arrays.

query.fetch1 is used when query contains exactly one entity, otherwise fetch1 will raise an error.

query.fetchn returns an arbitrary number of elements with character arrays and blobs returned in the form of cell arrays, even when query happens to contain a single entity.

% when tab has exactly one entity:
[name, img] = query.fetch1('name', 'image');

% when tab has any number of entities:
[names, imgs] = query.fetchn('name', 'image');

Obtaining the primary key along with individual values

It is often convenient to know the primary key values corresponding to attribute values retrieved by fetchn. This can be done by adding a special input argument indicating the request and another output argument to receive the key values:

% retrieve names, images, and corresponding primary key values:
[names, imgs, keys] = query.fetchn('name', 'image', 'KEY');

The resulting value of keys will be a column array of type struct. This mechanism is only implemented for fetchn.

Rename and calculate

In DataJoint for MATLAB, all fetch methods have all the same capability as the proj operator. For example, renaming an attribute can be accomplished using the syntax below.

[names, BMIs] = query.fetchn('name', 'weight/height/height -> bmi');

See Proj for an in-depth description of projection.

Sorting and limiting the results

To sort the result, add the additional ORDER BY argument in fetch and fetchn methods as the last argument.

% retrieve field ``course_name`` from courses
% in the biology department, sorted by course number
notes = fetchn(university.Course & 'dept="BIOL"', 'course_name', ...
     'ORDER BY course');

The ORDER BY argument is passed directly to SQL and follows the same syntax as the ORDER BY clause

Similarly, the LIMIT and OFFSET clauses can be used to limit the result to a subset of entities. For example, to return the most advanced courses, one could do the following:

s = fetch(university.Course, '*', 'ORDER BY course DESC LIMIT 5')

The limit clause is passed directly to SQL and follows the same rules

Talk to the Community