I maintain several Ubuntu systems and needed a simple bash script that would backup / mirror these machines. Google pointed me to rsync. This article describes what I did with it. * mirror gist * rsync man page Synopsis remote_mirror is a bash function that calls rsync under the hood. For example,
This post follows Golub and Van Loan, introducing Householder relections and Givens rotations, then using these tools to sketch out implementations of QR, Hessenberg, and Schur decompositions. import numpy as np np.set_printoptions( precision=3, linewidth = 132, suppress = True ) First, create some test data: # Set dimensions and create identity
$$ \text{global} \, \LaTeX \newcommand{\matr}[1]{\mathbf{#1}} \newcommand{\tran}{^{\mkern-1.5mu\mathsf{T}}} \newcommand{\cond}[2] { \left. #1 \, \right\vert #2 } \DeclareMathOperator{\ProbOp}{P} \newcommand{\Pr}[1] { \ProbOp #1 } \newcommand{\Prp}[1] { \ProbOp \left( #1 \right) } \newcommand{\Prs}[1] { \ProbOp \left[ #1 \right] } \newcommand{\Prb}[1] { \ProbOp \left\{ #1
Abstract A homogeneous Poisson process is the simplest way to describe events that arrive in time. Here, we are interested in a collection of realizations. An example is user transactions in a system. Over time, we expect each user to produce a sequence of transaction events, and we would like
Abstract Censored data is an artifact of partial or incomplete measurements. A typical scenario would be a survival analysis of time to event data. For example, a study may end before a final measurement is available (right censoring). Another situation might occur when batch processing log file data: the reported
Synopsis I recently needed to merge the glyphs in two TrueType font files. FontForge, in particular its python extension, appeared to be the tool for the job. The example below follows the old FontForge documentation. Newer documentation is available here. The “Barebones” Solution Pretty simple: sit in a loop and
$$ \DeclareMathOperator*{\min}{min} $$ This work was originally published as an Inferentialist blog post. Synopsis The probability integral transform is a fundamental concept in statistics that connects the cumulative distribution function, the quantile function, and the uniform distribution. We motivate the need for a generalized inverse of the CDF and prove
This work was originally published as an Inferentialist blog post. Abstract Adaptive rejection sampling is a statistical algorithm for generating samples from a univariate, log-concave density. Because of the adaptive nature of the algorithm, rejection rates are often very low. The exposition of this algorithm follows the example given in
This work was originally published as an Inferentialist blog post. Synopsis An analysis done as part of a recent Uber interview, this post showcases a regularized logistic regression model used to assess customer retention. Background This is the Uber Interview Challenge, a screening test for potential data science interview candidates.
This work was originally published as an Inferentialist blog post. Introduction Last week, I went looking for a solution to the following problem: I have a single Linux server with a dynamically assigned IP address and I want to host several sites on this server. My registrar is Namecheap.com,
This work was originally published as an Inferentialist white paper. Abstract Lending Club is a relatively new, and fastly growing, peer to peer lending platform. Using historical data provided by the company, we describe a method for constructing optimial portfolios of Lending Club loans. Our results, driven by expected returns,
$$ \DeclareMathOperator*{\argmax}{arg\,max} \DeclareMathOperator*{\Bern}{Bernoulii} \DeclareMathOperator*{\Bin}{Binomial} $$ This work was originally published as an Inferentialist white paper. Introduction A friend, also an employee at a large, Seattle-area company, recently approached me with the following problem. Suppose we wanted to oversubscribe the shared resources that we lease to