When runnin plot() function I get the error below even though matplotlib is installed. Also the algorithm is pretty slow when running on something bigger than toy example.

Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "/usr/local/lib/python2.7/dist-packages/physt/plotting/__init__.py", line 137, in __call__
    return plot(self.histogram, kind=kind, **kwargs)
  File "/usr/local/lib/python2.7/dist-packages/physt/plotting/__init__.py", line 91, in plot
    backend_name, backend = _get_backend(backend)
  File "/usr/local/lib/python2.7/dist-packages/physt/plotting/__init__.py", line 70, in _get_backend
    raise RuntimeError("No plotting backend available. Please, install matplotlib (preferred) or bokeh (limited).")
RuntimeError: No plotting backend available. Please, install matplotlib (preferred) or bokeh (limited).

Thanks for writing this awesome library!

I have a question regarding smoothing of polar 2D histograms. I am constructing a histogram like described on this page https://physt.readthedocs.io/en/latest/special_histograms.html#Polar-histogram and now I want to smooth it with a Gaussian kernel (like scipy.ndimage.gaussian_filter). What is the most elegant / correct method to do that?

Hi I found a project on rebinning histogram at https://github.com/jhykes/rebin and I opened an issue (jhykes/rebin#5) on that project page asking about integrating his code to this project. I hope you will appreciate it.

If you have a large dataset with a small number of values (such as consisting only of integers 1-10) then it would be nice to have the bin x-axis labels at the center under the respective bin instead of at the bin edges.

I recognise this case is more of a 'histogram as bar plot' kind of thing, but it is a use-case I have often.

Hi, I have tried the example of spherical histogram. After a small modification of the code (normalized the data as unit vectors),

n = 100 data = np.empty((n, 3)) data[:,0] = np.random.normal(0, 1, n) data[:,1] = np.random.normal(0, 1, n) data[:,2] = np.random.normal(0, 1, n) for i in range(n): scale = np.sqrt(data[i,0]**2 + data[i,1]**2 + data[i,2]**2) data[i,0] = data[i,0]/scale data[i,1] = data[i,1]/scale data[i,2] = data[i,2]/scale

h = special.spherical_histogram(data, theta_bins=20, phi_bins=20) ax.scatter(data[:,0], data[:,1], data[:,2])

globe = h.projection("theta", "phi") globe.plot.globe_map(density=True, figsize=(7, 7), cmap="rainbow")

plt.show()

I got an error: “RuntimeError: Bins not in rising order.” What did I do wrong? Thank you for your support.

I'm following the paper (http://jmlr.org/papers/volume11/ben-haim10a/ben-haim10a.pdf) implemented by https://github.com/carsonfarmer/streamhist, and the notion of approximate histograms seems elegant and efficient.

After seeing the internals of streamhist (trying to fix bugs) and reading the paper, I can imagine ways to make a better implementation: e.g. much more efficient discovery of bins to be joined, and avoiding temporary lists when possible. Also the code seems overly complex, partially due to features like "bin freezing" which try to workaround poor bin joining performance.

Anyway since streamhist is defunct, I'm thinking about trying an implementation. I wonder if this kind of histogram would fit into physt (and if sortedcollections would be reasonable as a dependency).

name: physt (?) github tag line: P(i/y)thon h(i/y)stograms (???)

google search for "python streaming histogram"

• top result is https://github.com/carsonfarmer/streamhist (unused / unmaintained)
• physt not in initial 10 pages of results...

For over a year I've wanted to find a Python library which supports efficient histogram updates without a bunch of ugly dependencies. I've searched many times. Today I happened to get lucky by seeing physt mentioned at the bottom of a SO question (https://stackoverflow.com/questions/40627274/).

To improve discoverability by search, please consider updating the github tag line to concisely and accurately describe the library (... rather than be cute).

If you try to merge bins:

from physt import h2
from scipy.stats import multivariate_normal
hist = h2(*multivariate_normal.rvs((0,0), size=100_000).T, bins=100)
hist.merge_bins(2)

You get a warning from numpy:

/home/schreihf/.local/lib/python3.7/site-packages/physt/histogram_base.py:572: FutureWarning: Using a non-tuple sequence for multidimensional indexing is deprecated; use `arr[tuple(seq)]` instead of `arr[seq]`. In the future this will be interpreted as an array index, `arr[np.array(seq)]`, which will result either in an error or a different result.
  new_frequencies[new_index] += old_frequencies[old_index]
/home/schreihf/.local/lib/python3.7/site-packages/physt/histogram_base.py:573: FutureWarning: Using a non-tuple sequence for multidimensional indexing is deprecated; use `arr[tuple(seq)]` instead of `arr[seq]`. In the future this will be interpreted as an array index, `arr[np.array(seq)]`, which will result either in an error or a different result.
  new_errors2[new_index] += old_errors2[old_index]

• [x] Analogous data model to Histogram1D
• [x] refactor HistogramBase class -> common behaviour of 1D and 2D
• [x] revisit binning schemas
• [x] histogram2D facade function to be compatible with numpy one
• [x] plotting
• [x] arithmetic operations
• [x] documentation
• [ ] stats

[SOMEDIR}\physt\physt\plotting\plotly.py in <module>
     12 
     13 import plotly.offline as pyo
---> 14 import plotly.plotly as pyp
     15 import plotly.graph_objs as go
     16 

~\Miniconda3\lib\site-packages\plotly\plotly\__init__.py in <module>
      2 from _plotly_future_ import _chart_studio_error
      3 
----> 4 _chart_studio_error("plotly")

~\Miniconda3\lib\site-packages\_plotly_future_\__init__.py in _chart_studio_error(submodule)
     41 
     42 def _chart_studio_error(submodule):
---> 43     raise ImportError(
     44         """
     45 The plotly.{submodule} module is deprecated,

ImportError: 
The plotly.plotly module is deprecated,
please install the chart-studio package and use the
chart_studio.plotly module instead. 

I have found that the bars in polar_map are centered on the left edge of the phi bins instead of their center. Because of this, the representation of the histogram does not coincide with the data, as in the figure below:

I think this can be easily solved by replacing

bars = ax.bar(phipos[i], dr[i], width=dphi[i], bottom=rpos[i], color=bin_color,

with

bars = ax.bar(phipos[i] + 0.5*dphi[i], dr[i], width=dphi[i], bottom=rpos[i], color=bin_color,

in the definition of polar_map.

By the way, thank you for this amazing package!

If the computed range for the binning divided by the number of bins is lower than the minimum float difference at the scale, we receive an error [ValueError: Bins not in rising order.] which is not very informative.

To reproduce:

data = [1, np.nextafter(1, 2)]
physt.h1(data)

It also happens when the range is 0, like in:

data = [1, 1]
physt.h1(data)