Permalink
Commits on Jun 26, 2017
  1. Merge pull request #1447 from schofiaj/patch-7

    peircej committed Jun 26, 2017
    ENH: Integrates 'coloured' noise into filtered noise
  2. Merge pull request #1448 from jonathanoroberts/master

    peircej committed Jun 26, 2017
    add seeding capability to psychojs random loops
Commits on Jun 23, 2017
  1. ENH: Integrates 'coloured' noise into filtered noise

    schofiaj committed Jun 23, 2017
    Modification to recently added noise stimulus class to remove ambigous 'coloured' noise term and integrate the desired fucntionality as a skew to the spectrum used as the starting point for filtered noise. Setting NoiseFilerOrder to 0 will turn off the filter leaving just the skew (just another kind of filtering after all) setting noiseFractalPower to 0 will produce an flat inital spectrum. 'White' noise is retained as a fast route to unfiltered spectra.
  2. RF: psychojs.zip should not be included in the repository

    peircej committed Jun 23, 2017
    It' something that gets compiled from the folder at build but it's just
    a zipped copy of the psychojs submodule and doesn't need duplicating
    
    [ci skip]
  3. Merge pull request #1446 from theodleif/master

    peircej committed Jun 23, 2017
    RF: getting ioHub working under linux again
  4. Merge pull request #1445 from kaczmarj/fix-setup-py3

    peircej committed Jun 23, 2017
    FIX: pip install on python3
  5. re-switched off debug info

    theodleif committed Jun 23, 2017
Commits on Jun 22, 2017
  1. add seeding capability to psychojs random loops

    jonathanoroberts committed Jun 22, 2017
    This adds a line to the default index.html template for web
    experiments. See pull request with same title in the psychojs
    repository for changes to the psychojs code to implement this.
    
    This uses seedrandom() from https://github.com/davidbau/seedrandom
    and will require adding the seedrandom.min.js from that repository
    to the psychojs/js/vendors/ folder in the psychojs.zip file
    that is packaged with the psychopy standalone app.
  2. Merge pull request #1444 from peircej/master

    peircej committed Jun 22, 2017
    RF: mostly fixes for Py3 compatibility and new dependency auto-install
  3. DOCS: updated requirements files

    peircej committed Jun 22, 2017
    You should now be able to install psychopy and its basic requirements using
    	pip install psychopy
    or
    	pip install -r psychopy
    
    The main requirements.txt now uses the setup.py requirements by doing a
    simple -e . installation. See
    	https://caremad.io/posts/2013/07/setup-vs-requirement/
    
    One advantage of this approach is that we can easily make requirements
    system-dependent (e.g. only add pyqt5 for Py3, and objc-core for darwin)
  4. DOCS: remove redundant message on win32 installer

    peircej committed Jun 22, 2017
    Used to warn people not to start the install until uninstall was
    complete, but that's no longer possible so ermove the message
Commits on Jun 20, 2017
  1. Merge pull request #1442 from schofiaj/patch-6

    peircej committed Jun 20, 2017
    ENH: Stimulus class for generating random noise stimuli
  2. ENH: save monitor calibrations as JSON files

    peircej committed Jun 20, 2017
    Currently:
    	- this is *in addition* to pickle files (.calib)
    	- only pickle files are loaded
    Next step (to be done immediately):
    	- load the JSON file in preference if it exists
    3rd step (to be done in the future, with no rush):
    	- stop saving pickle files altogether
Commits on Jun 16, 2017
  1. ENH: Stimulus class for generating random noise stimuli

    schofiaj committed Jun 16, 2017
    I offer this stimulus class for generating visual random noise stimuli of various kinds. There are 4 main classes. Pixel based, Gabor element based, Filtered noise, and Coloured noise. Pixel based can have grey levels distributed according to binary, normal of uniform distributions. Gabor based can be oriented or isotropic. Filter based used Butterworth filtered white noise. Coloured noise allows gives the noise a skewed spatial frequency spectrum and includes white noise and noise derived from the spectrum of an image - this tool randomises the phase spectrum of the image. I’m not 100% sure about using the term coloured noise. I’ve used it as the technical terms for not flat spectra mostly use colour names: pink-noise being the most common. But is might be confused with noise that actually has different hues. I am working on a builder component to go with this. That component will be quite complex with a lot of parameters. I would like it for my lab but wonder how it might be simplified eg different builder components for different classes of noise or different tabs for the parameter belonging to different classes of noise. Do you have a view about adding such a component? I’m happy to upload what I have for discussion at some point soon.
  2. Merge pull request #1441 from peircej/master

    peircej committed Jun 16, 2017
    FF: mostly fixes for Py3 compatibility
  3. DOCS: version bump to 1.85.2

    peircej committed Jun 15, 2017
  4. DOCS: updated changelog

    peircej committed Jun 16, 2017
    # Conflicts:
    #	psychopy/CHANGELOG.txt
  5. FF: only perform a decode it the filename is a bytes object

    peircej committed Jun 16, 2017
    On Py2 it was always a str/bytes obj but onPy3 it is more likely a
    str/unicode
  6. BF: return focus to the main window after closing tools>update dialog

    peircej committed Jun 16, 2017
    The problem was that we were doing Destroy() rather than Close() on
    the dialog