-
The optimize functions do not have any way to set parameter limits, for example stating p[0]>=0 to prevent the fitting from attempting negative values. This SEVERALLY limits the usefulness of the pack…
-
To focus the [conversation](https://github.com/dotnet/roslyn/issues/11324) with @CyrusNajmabadi about [SourceGenerators](https://github.com/dotnet/roslyn/blob/master/docs/features/generators.md) and m…
-
release condition: (spyder3.0.2, numpy-1.11.2, pandas 0.19.1, scikit-learn-0.18.0, scikit-optimize-0.2, Matplotlib1.5.3 (rc3: november 5th 2016)
Features:
- Highlights : pandas 0.19.1, scikit-lear…
-
there is an issue to install packages with entry points, not sure if it's a 3.6.0b1 bug or a feature. python3.6 may look for a registery entry, and react badly when not finding one.
to read: sys.path…
-
release condition: (Ipython 5.1, PyQt5.7, Scipy 0.18.0, Bokeh 0.12.1), (release: august 28th 2016)
Features:
Highlights :
- user facing : spyder3.0b4+, Bokeh 0.12.1
- computation: dask/distributed/x…
-
There are many scenarios where you'd like to group a set of typed values temporarily, without the grouping itself warranting a "concept" or type name of its own.
Other languages use variations over t…
-
I think NLopt is probably the right choice. Here is a note on this from @hongkai-dai:
I did a quick search, and it seems IPOPT is a good choice. Both IPOPT and and NLopt are free and open source. IP…
-
Tweedie is also in the linear exponential family and is use, for example, to model claim amounts.
Tweedie is a special case of compound Poisson distribution, that has the pdf/likelihood specified as …
-
I have used scipy.curve_fit before and was able to fit my data to the following non_linear function. I don't remember the older version number.
However, now I am trying to fit the curve on the same …
-
I was adding a QuantileLoss type when I noticed this. In statistics, `residual = y - yhat`. Our distance losses are implemented as if `residual = yhat - y`, so derivatives are missing a minus sign.
…