Gridfix

**Bugfixes**
- Fixed error when mapping a dataset without fixation times (2)

New Features
- gridfix is now available via pip (PyPI)!
- Add number of viewing passes and refixations as Dependent Variables

This release also tags the specific version used for the results of an upcoming manuscript.

New Features:
- Add support for object bounding box masks (BBoxRegionSet)
- Add fixation duration-based DVs (first fixation, single fixation, total gaze, time to initial fixation)
- Support exclusion of the last fixated region per image, in addition to first
- Support mapping of fixations outside of all explicit regions (background fixations)
- Add fixation heat maps

Bugfixes:
- Correct default parameter value for CentralBiasFeature
- Plotting fixes and improvements
- Add missing LICENSE file (gridfix is MIT licensed)

This is the first release of GridFix after acceptance of the corresponding manuscript (see https://www.frontiersin.org/articles/10.3389/fnhum.2017.00491/abstract).

Supported features:
- Image parcellations based on regular grid
- Explicit modeling of Central Viewer Bias (euclidean distance, Gaussian map, taxicab/manhattan distance)
- Low-level image features: luminance, luminance contrast, edges
- Feature values from saliency maps (mean, peak, ...)
- Generation of predictor files and exemplary R source code for GLMM analysis

Please refer to the documentation for usage examples and references:
https://ischtz.github.io/gridfix/