The MongoAdapter module reads data from a Mongo database collection and produces a NumPy array containing the loaded. The following features are currently implemented:

  • The MongoAdapter engine is written in C to ensure data is loaded fast with minimal memory usage.
  • Python slicing notation can be used to specify the subset of records to be read from the data source.
  • The MongoAdapter engine has automatic type inference so the user does not have to specify dtypes of the output array.


The MongoAdapter module contains the follwowing constructor for creating MongoAdapter objects:

MongoAdapter (host, port, database, collection)
MongoAdapter contructor
host - Host name where Mongo database is running.
port - Port number where Mongo database is running.
database - Mongo database to connect to
collection - Mongo database collection
set_field_names (names)
Set field names to read when creating output NumPy array.
get_field_names ()
Returns names of fields that will be read when reading data from Mongo database.
set_field_types (types=None)
Set NumPy dtypes for each field, specified as a dict of field names/indices and associated dtype. (Example: {0:’u4’, 1:’f8’, 2:’S10’})
get_field_types ()
Returns dict of field names/indices and associated NumPy dtype.

The MongoAdapter object contains the following properties:

size (readonly)
Number of documents in the Mongo database + collection specified in constructor.

Basic Usage

  1. Create MongoAdapter object for data source

    >>> import iopro
    >>> adapter = iopro.MongoAdapter('localhost', 27017, 'database_name', 'collection_name')
  2. Load Mongo collection documents into NumPy array using slicing notation

    >>> # read all records for 'field0' field
    >>> array = adapter['field0'][:]
    >>> # read first ten records for 'field0' and 'field1' fields
    >>> array = adapter[['field0', 'field1']][0:10]
    >>> # read last record
    >>> array = adapter['field0'][-1]
    >>> # read every other record
    >>> array = adapter['field0'][::2]