scidb-py 18.1.1

Python library for SciDB

Warning

Since SciDB-Py Release 16.9.1 (released in September 2017) the library has been rewritten entirely from scratch. 16.9.1 and newer versions are not compatible with the previous versions of the library. The documentation for the previous versions is available at SciDB-Py documentation (legacy). GitHub pull requests are still accepted for the previous versions, but the code is not actively maintained.

Version Information

The major and minor version numbers of SciDB-Py track the major and minor version of SciDB they are compatible with. For example SciDB-Py 16.9.1, 16.9.2 or 16.9.10 are all compatible with SciDB 16.9.x.

During SciDB 16.9, Shim (HTTP service for SciDB) transitioned from query authentication to session authentication. SciDB-Py has been updated to be compatible with the new Shim. Below is the compatibility matrix between SciDB-Py and Shim:

SciDB-Py	Shim
16.9.1	query authentication (old Shim)
16.9.2	query authentication (old Shim)
16.9.10	session authentication (new Shim)

From 16.9.10 onwards only Shim with session authentication is supported.

Requirements

SciDB 16.9 or newer with Shim

Python 2.7.x, 3.4.x, 3.5.x, 3.6.x or newer.

Required Python packages:

backports.weakref
enum34
numpy
pandas
requests
six

CentOS 6 and Red Hat Enterprise Linux 6

CentOS 6 and Red Hat Enterprise Linux 6 come with Python 2.6. SciDB-Py requires Python 2.7 or newer (see above). The default Python cannot be upgraded on these operating systems. Instead a different Python version can be installed in parallel using Software Collections. For example, here are the instructions to install Python 2.7 using Software Collections.

Installation

Install latest release:

$ pip install scidb-py

Install development version from GitHub:

$ pip install git+http://github.com/paradigm4/scidb-py.git

Documentation

See SciDB-Py Documentation.