=========

• Python 3 compatible - installation via pip more straightforward
• New command-line interface - separate analysis scripts merged into improved command-line structure
• Improved documentation - access hints about options directly on the command line (type colonyzer -h at command prompt)
• Easier grid location - only specify gridded array format, Colonyzer.txt calibration files should not be necessary for most images
• Improved grid location algorithms - estimate grid width by autocorrelation analysis, grid location and rotation by gradient-based optimisation (with optional global search)
• Improved spot location algorithm - recursive estimation of centre of mass of each spot, subject to no increase in signal along tile edge, refines grid location estimates for irregular spots
• Alternative segmentation algorithm - detects cells by first detecting culture edges and infilling: even higher sensitivity at very low signal levels
• Improved lighting correction - generate pseudo-empty plate by infilling segmented pixels using a Gaussian Random Markov Field update
• Test datasets - suite of problematic images for checking performance of scripts and algorithms

First, increment the version number in setup.py.

If you have not already registered a username with PYPI you should do that. Then authenticate yourself:

python setup.py register

Then, navigate to the directory containing setup.py and upload the updated package at the command-line as follows:

python setup.py sdist upload

Website describing current, stable release: http://research.ncl.ac.uk/colonyzer/

Open access manuscript describing Colonyzer algorithms: http://dx.doi.org/10.1186/1471-2105-11-287

Open access video and manuscript demonstrating the use of Colonyzer within a Quantitative Fitness Analysis workflow http://www.jove.com/video/4018