The softmax exp(x)/sum(exp(x)) is actually numerically well-behaved. It has only positive terms, so we needn’t worry about loss of significance, and the denominator is at least as large as the numerator, so the result is guaranteed to fall between 0 and 1. The only accident that might happen is over- or under-flow in the exponentials. … Read more
I am also in academia and I have written quantum mechanical simulation programs to be executed on our cluster. I made the same observation regarding testing or even version control. I was even worse: in my case I am using a C++ library for my simulations and the code I got from others was pure … Read more
At some stage you’re going to need floating point arithmetic. It’s hard to do it well, less hard to do it competently, and easy to do it badly. This paper is a must read: What Every Computer Scientist Should Know About Floating-Point Arithmetic
I already posted this in another thread but I think it fits better in this one: UPDATE (30.07.2014): I re-run the the benchmark on our new HPC. Both the hardware as well as the software stack changed from the setup in the original answer. I put the results in a google spreadsheet (contains also the … Read more
Yes, the same order is guaranteed if it is not modified. See the docs here. Edit: Regarding if changing the value (but not adding/removing a key) will affect the order, this is what the comments in the C-source says: /* CAUTION: PyDict_SetItem() must guarantee that it won’t resize the * dictionary if it’s merely replacing … Read more
There is OxyPlot which I recommend. It has packages for WPF, Metro, Silverlight, Windows Forms, Avalonia UI, XWT. Besides graphics it can export to SVG, PDF, Open XML, etc. And it even supports Mono and Xamarin for Android and iOS. It is actively developed too. There is also a new (at least for me) open … Read more
Most sklearn objects work with pandas dataframes just fine, would something like this work for you? import pandas as pd import numpy as np from sklearn.decomposition import PCA df = pd.DataFrame(data=np.random.normal(0, 1, (20, 10))) pca = PCA(n_components=5) pca.fit(df) You can access the components themselves with pca.components_
import random def decision(probability): return random.random() < probability
I am curious as to how F# performance compares to C++ performance? Varies wildly depending upon the application. If you are making extensive use of sophisticated data structures in a multi-threaded program then F# is likely to be a big win. If most of your time is spent in tight numerical loops mutating arrays then … Read more
What languages/environments have you used for developing scientific software, esp. data analysis? What libraries? (E.g., what do you use for plotting?) I used to work for Enthought, the primary corporate sponsor of SciPy. We collaborated with scientists from the companies that contracted Enthought for custom software development. Python/SciPy seemed to be a comfortable environment for … Read more