lebigot commented on issue lebigot/uncertainties#30
@lebigot

PS: ChainMaps actually do not solve the quadratic time problem, because the lookup of shared variables between terms would become linear, for each …

@lebigot
Chinese characters cannot be entered
lebigot commented on issue lebigot/uncertainties#30
@lebigot

Note to self: the "lazy evaluation of derivatives via formal expression" idea above does not solve the problem of the slowness of sum(). In fact, c…

@lebigot
lebigot closed pull request lebigot/Skyloop#1
@lebigot
Update to final version
lebigot opened pull request lebigot/Skyloop#1
@lebigot
Update to final version
4 commits with 274,108 additions and 17,356 deletions
@lebigot
  • @lebigot 5c205b4
    Colors on the Python Package Index page.
lebigot commented on issue lebigot/uncertainties#30
@lebigot

Note to self: I kept thinking about an optimization for sums, that could also work in the general case. I came up with the following principles: I…

lebigot commented on issue lebigot/uncertainties#30
@lebigot

Note to self: even multiplication could in principle be accelerated similarly, with a generalization of the ChainMap method: when multiplying to va…

lebigot commented on issue lebigot/uncertainties#30
@lebigot

I think I know how summing could be sped up: by using the collections.ChainMap class in order to store the derivatives of the sum, based on the der…

lebigot commented on issue lebigot/uncertainties#30
@lebigot

I had a look. What you observe is normal: every time one term is added, the program calculates the derivatives of the new sum with respect to all t…

@lebigot
  • @lebigot a5719f9
    Added comment about Python 2.7+ syntax.
lebigot commented on issue lebigot/uncertainties#30
@lebigot

Wow, thanks for all these tests. I'll have a close look in the coming days.

lebigot commented on issue lebigot/uncertainties#30
@lebigot

I have not looked at this yet (having two jobs takes a toll!). Now, this looks like an issue that's probably concentrated on a small part of the co…

lebigot opened issue numpy/numpy#5927
@lebigot
Incomplete `order` argument documentation
lebigot commented on pull request astropy/astropy#3715
@lebigot

Good that you welcome further thoughts: I have been needed mean() etc. with uncertainties quite a few times. Handling its uncertainty is almost OK:…

lebigot commented on pull request astropy/astropy#3715
@lebigot

Ah, sorry, I had missed the fact that you had put np.cbrt(x): I thought it was only x. Good timing results (I guess that in your example you have 2…

@lebigot
Data types not inferred
lebigot commented on issue lebigot/uncertainties#19
@lebigot

As far as I can tell, yes, it is similar to what you were proposing, but I haven't checked the details.

lebigot commented on pull request astropy/astropy#3715
@lebigot

Shouldn't this be **(2/3)? (__future__ is used, so no need for 2./3).

lebigot commented on pull request astropy/astropy#3715
@lebigot

Thanks @mhvk: I will have a look at what you did. The "big outstanding issue" that you are mentioning (large memory consumption for even simple arr…

lebigot commented on issue lebigot/uncertainties#19
@lebigot

For reference: direct implementation of uncertainties in NumPy ndarrays: astropy/astropy#3715 (comment).

lebigot commented on issue databricks/spark-csv#59
@lebigot

Thanks, this was very helpful: this was close enough to the answer to my need. :) I wanted to use spark-csv I would strongly suggest that you updat…

@lebigot
Add simple install instructions details