Release v0.00.21 - PyTorch histogramming + interactive error bars

PyTorch histogramming

• considered to be ported directly to PyTorch
• Syntax as in the nupmy histogramdd
  • except of the normed option
  • support for GPU
  • summary perfromance comparion:
    • https://docs.google.com/presentation/d/1vYPTTsn5Zs7X9OWaewUTlnw9hqJ3XKJNGKiAViHv-cY
• Example/test usage:
• Benchmark:

Interactive error bars

• working for pandas
• for ROOT trees additional parsing to be done
• see example RootInteractive/InteractiveDrawing/bokeh/test_bokehDrawSA.py
• Example syntax:

figureArray = [
#   ['A'], ['C-A'], {"color": "red", "size": 7, "colorZvar":"C", "filter": "A<0.5"}],
    [['A'], ['C-A'], {"color": "red", "size": 7, "colorZvar": "C", "errY": "errY", "errX":"0.01" }],
    [['A'], ['C+A', 'C-A']],
    [['B'], ['C+B', 'C-B'], {"color": "red", "size": 7, "colorZvar": "C", "errY": "errY" }],
    [['D'], ['(A+B+C)*D'], {"size": 10, "errY": "errY"} ],
    [['D'], ['D*10'], {"size": 10, "errY": "errY"}],
]