In scientific programming speed is important. Functions written for general public use have a lot of control-flow checks which are not necessary if you are confident enough with your code.To quicken your code execution I suggest to strip run-of-the-mill functions to their bare bones. You can save serious wall-clock time by using only the laborers code lines. Below is a walk-through example of what I mean.

# Category: Code

R-code

## Publication in Significance – code

Couple of months ago I published a paper in Significance – couple of pages describing the essence of deep learning algorithms, and why they are so popular. I got a few requests for the code which generated the figures in that paper. This weekend I reviewed my code and was content to see that I used a pseudorandom numbers, with a seed (as oppose to completely random numbers; without a seed). So now the figures are exactly reproducible. The actual code to produce the figures, and the figures themselves (e.g. for teaching purposes) are provided below.

## R tips and tricks – Timing and profiling code

Modern statistical methods use simulations; generating different scenarios and repeating those thousands of times over. Therefore, even trivial operations burden computational speed.

In the words of my favorite statistician Bradley Efron:

“There is some sort of law working here, whereby statistical methodology always expands to strain the current limits of computation.”

In addition to the need for faster computation, the richness of open-source ecosystem means that you often encounter different functions doing the same thing, sometimes even under the same name. This post explains how to measure the computational efficacy of a function so you know which one to use, with a couple of actual examples for reducing computational time.

## R + Python = Rython

Enough! Enough with that pointless R versus Python debate. I find it almost as pointless as the Bayesian vs Frequentist “dispute”. I advocate here what I advocated there (“..don’t be a Bayesian, nor be a Frequenist, **be opportunist**“).

Nowadays even marginally tedious computation is being sent to faster, minimum-overhead languages like C++. So it’s mainly syntax administration we insist to insist on. What does it matter if we have this:

1 2 3 |
xsquare <- function(x){ x^2 } |

Or that

1 2 3 4 |
def xsquare(x): return x**2 |

## R Journal publication

The R Journal is the open access, refereed journal of the R project for statistical computing. It features short to medium length articles covering topics that should be of interest to users or developers of R.

Christoph Weiss, Gernot Roetzer and myself have joined forces to write an R package and the accompanied paper: **Forecast Combinations in R using the ForecastComb Package**, which is now published in the R journal. Below you can find a few of my thoughts about the journey towards publication in the R journal, and a few words about working with a small team of three, from three different locations.

## R tips and tricks – the pipe operator

The R language has improved over the years. Amidst numerous splendid augmentations, the `magrittr`

package by Stefan Milton Bache allows us to write more readable code. It uses an ingenious piping convention which will be explained shortly. This post talks about when to use those pipes, and when to avoid using pipes in your code. I am all about ~~that bass~~ readability, but I am also about speed. Use the pipe operator, but watch the tradeoff.

## R vs MATLAB – round 4

This is another comparison between R and MATLAB (Python also in the mix this time). In previous rounds we discussed the differences in 3d visualization, differences in syntax and input-output differences. Today is about computational speed.

## R tips and tricks – Package Dependencies

In this post about the most popular machine learning R packages I showed the incredible- exponential growth displayed by R software, measured by the number of package downloads. Here is another graph which shows a more linear growth in R (and an impressive growth in python) as measured by % of question posted in stack overflow

## R tips and tricks – Set Working Directory

This is more an Rstudio tip than an R tip. It would be nice to know how the following works for different editors, but Rstudio is common enough and awesome enough for the following to be relevant.

## R tips and tricks – Faster Loops

## Insert or bind?

This is the first in a series of planned posts, sharing some R tips and tricks. I hope to cover topics which are not easily found elsewhere. This post has to do with loops in R. There are two ways to save values when looping:

1. You can predefine a vector and fill it, or

2. you can recursively bind the values.

Which one is faster?

## Good coding practices – part 2

## Introduction

In part 1 of *Good coding practices* we considered how best to code for someone else, may it be a colleague who is coming from Excel environment and is unfamiliar with scripting, a collaborator, a client or the future-you, the you few months from now. In this second part, I give some of my thoughts on how best to write functions, the do’s and dont’s.

## Good coding practices – part 1

## Introduction

At work, I recently spent a lot of time coding for someone else, and like anything else you do, there is much to learn from it. It also got me thinking about scripting, and how best to go about it. To me it seems that the new working generation mostly tries to escape from working with Excel, but “let’s not kid ourselves: the most widely used piece of software for statistics is Excel” (Brian D. Ripley). this quote is 15 years old almost, but Excel still has a strong hold on the industry.

Here I discuss few good coding practices. Coding for someone else is not to be taken literally here. ‘Someone else’ is not necessarily a colleague, it could just as easily be the “future you”, the you reading your code six months from now (if you are lucky to get responsive referees). Did it never happened to you that your past-self was unduly cruel to your future-self? that you went back to some old code snippets and dearly regretted not adding few comments here and there? Of course it did.

Unlike the usual metric on which “good” is usually measured by when it comes to coding: *good = efficient*, here the metric would be different: *good = friendly*. They call this literate programming. There is a fairly deep discussion about this paradigm by John D. cook (follow what he has to say if you are not yet doing it, there is something for everyone).

## Forecast combinations in R

Few weeks back I gave a talk in the R/Finance 2016 conference, about forecast combinations in R. Here are the slides:

## Most popular machine learning R packages

The good thing about using open-source software is the community around it. There are very many R packages online, and recently CRAN package download logs were released. This means we can have a look at the number of downloads for each package, so to get a good feel for their relative popularity. I pulled the log files from the server and checked a few packages which are known to be related to machine learning. With this post you can see which are the community favorites, and get a feel for the R-software trend growth.

## Forecast averaging example

Especially in economics/econometrics, modellers do not believe their models reflect reality as it is. No, the yield curve does NOT follow a three factor Nelson-Siegel model, the relation between a stock and its underlying factors is NOT linear, and volatility does NOT follow a Garch(1,1) process, nor Garch(?,?) for that matter. We simply look at the world, and try to find an apt description of what we see.