nhaliday + tutorial   193

documentation - Materials for learning TikZ - TeX - LaTeX Stack Exchange
The way I learned all three was basically demand-driven --- "learning by doing". Whenever I needed something "new", I'd dig into the manual and try stuff until either it worked (not always most elegantly), or in desperation go to the examples website, or moan here on TeX-'n-Friends. Occasionally supplemented by trying to answer "challenging" questions here.

yeah I kinda figured that was the right approach. just not worth the time to be proactive.
q-n-a  stackex  latex  list  links  tutorial  guide  learning  yak-shaving  recommendations  programming  visuo  dataviz  prioritizing  technical-writing 
7 weeks ago by nhaliday
its-not-software - steveyegge2
You don't work in the software industry.


So what's the software industry, and how do we differ from it?

Well, the software industry is what you learn about in school, and it's what you probably did at your previous company. The software industry produces software that runs on customers' machines — that is, software intended to run on a machine over which you have no control.

So it includes pretty much everything that Microsoft does: Windows and every application you download for it, including your browser.

It also includes everything that runs in the browser, including Flash applications, Java applets, and plug-ins like Adobe's Acrobat Reader. Their deployment model is a little different from the "classic" deployment models, but it's still software that you package up and release to some unknown client box.



Our industry is so different from the software industry, and it's so important to draw a clear distinction, that it needs a new name. I'll call it Servware for now, lacking anything better. Hardware, firmware, software, servware. It fits well enough.

Servware is stuff that lives on your own servers. I call it "stuff" advisedly, since it's more than just software; it includes configuration, monitoring systems, data, documentation, and everything else you've got there, all acting in concert to produce some observable user experience on the other side of a network connection.
techtariat  sv  tech  rhetoric  essay  software  saas  devops  engineering  programming  contrarianism  list  top-n  best-practices  applicability-prereqs  desktop  flux-stasis  homo-hetero  trends  games  thinking  checklists  dbs  models  communication  tutorial  wiki  integration-extension  frameworks  api  whole-partial-many  metrics  retrofit  c(pp)  pls  code-dive  planning  working-stiff  composition-decomposition  libraries  conceptual-vocab  amazon  system-design  cracker-prog 
9 weeks ago by nhaliday
Use and Interpretation of LD Score Regression
LD Score regression distinguishes confounding from polygenicity in genome-wide association studies: https://sci-hub.bz/10.1038/ng.3211
- Po-Ru Loh, Nick Patterson, et al.


Both polygenicity (i.e. many small genetic effects) and confounding biases, such as cryptic relatedness and population stratification, can yield inflated distributions of test statistics in genome-wide association studies (GWAS). However, current methods cannot distinguish between inflation from bias and true signal from polygenicity. We have developed an approach that quantifies the contributions of each by examining the relationship between test statistics and linkage disequilibrium (LD). We term this approach LD Score regression. LD Score regression provides an upper bound on the contribution of confounding bias to the observed inflation in test statistics and can be used to estimate a more powerful correction factor than genomic control. We find strong evidence that polygenicity accounts for the majority of test statistic inflation in many GWAS of large sample size.

Supplementary Note: https://images.nature.com/original/nature-assets/ng/journal/v47/n3/extref/ng.3211-S1.pdf

An atlas of genetic correlations across human diseases
and traits: https://sci-hub.bz/10.1038/ng.3406


Supplementary Note: https://images.nature.com/original/nature-assets/ng/journal/v47/n11/extref/ng.3406-S1.pdf

ldsc is a command line tool for estimating heritability and genetic correlation from GWAS summary statistics. ldsc also computes LD Scores.
nibble  pdf  slides  talks  bio  biodet  genetics  genomics  GWAS  genetic-correlation  correlation  methodology  bioinformatics  concept  levers  🌞  tutorial  explanation  pop-structure  gene-drift  ideas  multi  study  org:nat  article  repo  software  tools  libraries  stats  hypothesis-testing  biases  confounding  gotchas  QTL  simulation  survey  preprint  population-genetics 
november 2017 by nhaliday
Anatomy of an SQL Index: What is an SQL Index
“An index makes the query fast” is the most basic explanation of an index I have ever seen. Although it describes the most important aspect of an index very well, it is—unfortunately—not sufficient for this book. This chapter describes the index structure in a less superficial way but doesn't dive too deeply into details. It provides just enough insight for one to understand the SQL performance aspects discussed throughout the book.

B-trees, etc.
techtariat  tutorial  explanation  performance  programming  engineering  dbs  trees  data-structures  nibble 
september 2017 by nhaliday
Introduction to Scaling Laws

Galileo’s Discovery of Scaling Laws: https://www.mtholyoke.edu/~mpeterso/classes/galileo/scaling8.pdf
Days 1 and 2 of Two New Sciences

An example of such an insight is “the surface of a small solid is comparatively greater than that of a large one” because the surface goes like the square of a linear dimension, but the volume goes like the cube.5 Thus as one scales down macroscopic objects, forces on their surfaces like viscous drag become relatively more important, and bulk forces like weight become relatively less important. Galileo uses this idea on the First Day in the context of resistance in free fall, as an explanation for why similar objects of different size do not fall exactly together, but the smaller one lags behind.
nibble  org:junk  exposition  lecture-notes  physics  mechanics  street-fighting  problem-solving  scale  magnitude  estimate  fermi  mental-math  calculation  nitty-gritty  multi  scitariat  org:bleg  lens  tutorial  guide  ground-up  tricki  skeleton  list  cheatsheet  identity  levers  hi-order-bits  yoga  metabuch  pdf  article  essay  history  early-modern  europe  the-great-west-whale  science  the-trenches  discovery  fluid  architecture  oceans  giants  tidbits  elegance 
august 2017 by nhaliday
Analysis of variance - Wikipedia
Analysis of variance (ANOVA) is a collection of statistical models used to analyze the differences among group means and their associated procedures (such as "variation" among and between groups), developed by statistician and evolutionary biologist Ronald Fisher. In the ANOVA setting, the observed variance in a particular variable is partitioned into components attributable to different sources of variation. In its simplest form, ANOVA provides a statistical test of whether or not the means of several groups are equal, and therefore generalizes the t-test to more than two groups. ANOVAs are useful for comparing (testing) three or more means (groups or variables) for statistical significance. It is conceptually similar to multiple two-sample t-tests, but is more conservative (results in less type I error) and is therefore suited to a wide range of practical problems.

good pic: https://en.wikipedia.org/wiki/Analysis_of_variance#Motivating_example

tutorial by Gelman: http://www.stat.columbia.edu/~gelman/research/published/econanova3.pdf

so one way to think of partitioning the variance:
y_ij = alpha_i + beta_j + eps_ij
Var(y_ij) = Var(alpha_i) + Var(beta_j) + Cov(alpha_i, beta_j) + Var(eps_ij)
and alpha_i, beta_j are independent, so Cov(alpha_i, beta_j) = 0

can you make this work w/ interaction effects?
data-science  stats  methodology  hypothesis-testing  variance-components  concept  conceptual-vocab  thinking  wiki  reference  nibble  multi  visualization  visual-understanding  pic  pdf  exposition  lecture-notes  gelman  scitariat  tutorial  acm  ground-up  yoga 
july 2017 by nhaliday
Latin spelling and pronunciation - Wikipedia
From Solodow's "Latin Alive": Classical Latin (for literature as opposed to common tongue) was formed out of crucible of nativist anxieties

The earliest continuous Latin texts we have date from the late third and early second centuries B.C.E., beginning with the comedies of Plautus. The Latin in these and the other texts that follow them for the next century displays a certain amount of variety, as we might expect: a large and expressive vocabulary, some freedom with genders, declensions, and conjugations, a certain diversity in inflections and syntax. But in the first half of the first century B.C.E.;., this changed quickly and definitively. A group of men set about to find and fix a suitable form for the language. Their goal was to settle the language once and for all, and, in an important sense, they succeeded. These men, of whom the two most familiar are Caesar (100-44 B.C.E.) and Cicero (106-43 B.C.E.), did not constitute an academy of the Latin language, like those established in modern times for French and Spanish. Instead, by their own conscious practice they shaped the language into a form that seemed pure and worthy.

Their concerted effort to give the Latin language a fixed form was driven in part by the linguistic unsettledness and disorder they perceived around them. Language - actual spoken language - perhaps always appears messy to the ears and eyes of some, but at that time and place the messiness may have been very marked. Rome from its beginnings had been a city of immigrants, and the conquests abroad and other social upheavals of the preceding century had brought into the capital a swarm of people who did not speak Latin as their native language or were not familiar with the variety characteristic of the city. Some men consequently feared the disappearance of authentic, correct Latin. In his history of Roman oratory, Cicero links the deplorable linguistic situation of his day with social changes: “In those days [a century earlier] nearly everybody who had lived in this city and not been corrupted by home-bred provincialism spoke correctly. But the passage of time unquestionably changed the situation for the worse, no less at Rome than in Greece. Many people from different places who spoke a debased language poured into Athens and into this city. The language therefore needs to be purified” (Brutus 258).

Another impetus was the recognition that the linguistic situation, if grave, was not irremediable. Here the model of the Greek language played an important part. As Cicero draws a parallel between the problems at Rome and those at Athens, so he and his contemporaries looked to the latter for guidance in finding a solution. The dialect of Athens, known as Attic, which had established itself among the various Greek dialects as the one most prestigious and most suitable for refined speech and writing, had itselfpassed through a period of conscious purification; this purified Attic Greek served the Romans as an example. And at the same time that Attic offered a model to imitate, Greek rhetoricians were extolling the virtues of language that was logical, unambiguous, and otherwise clear.

Goaded by the current unhappy state of Latin and drawn by a vision of how it might be bettered, Caesar, Cicero, and others set about the task of purifying Latin. They shunned rusticitas “rusticity,” anything that smacked of the countryside. They strove for urbanitas “urbanity, refinement,” and in the sphere of language this was synonymous with Latinitas “(genuine) Latin-ness”; this equation is evident in the passage quoted from Cicero, who identifies as the genuine and desirable variety of Latin the one that had been spoken in the city of Rome by native Romans.
language  foreign-lang  mediterranean  the-classics  wiki  reference  history  iron-age  medieval  early-modern  article  howto  tutorial  multi  twitter  social  commentary  quotes  gnon  unaffiliated  right-wing  statesmen  big-peeps  leadership  tribalism  us-them  migration  speaking  linguistics  quixotic  syntax  lexical 
june 2017 by nhaliday
:feed v1 - /fora/posts/~2017.4.12..21.14.00..fe17~
The goal of this demo was to show that building a Twitter replacement actually isn't that hard at all; and it can be done almost entirely on the frontend. As shown, you don't even have to use React/Redux. But that's probably the way to go if you want to build the real thing.
techtariat  urbit  software  decentralized  twitter  social  internet  web  programming  tutorial  project  gnon 
april 2017 by nhaliday
« earlier      
per page:    204080120160

bundles : initmeta

related tags

abstraction  academia  accretion  accuracy  acm  acmtariat  advanced  adversarial  advice  aggregator  ai  akrasia  algorithms  amazon  analysis  anglo  api  apollonian-dionysian  app  apple  applicability-prereqs  approximation  arbitrage  architecture  art  article  asia  assembly  atoms  attention  audio  bandits  bangbang  bare-hands  bayesian  ben-recht  benchmarks  berkeley  best-practices  better-explained  bias-variance  biases  big-peeps  big-picture  big-yud  bio  biodet  bioinformatics  biomechanics  bitcoin  blockchain  books  bootstraps  brands  build-packaging  business  c(pp)  caching  calculation  capital  career  carmack  CAS  causation  cheatsheet  checking  checklists  china  civic  cjones-like  classic  classification  clojure  cloud  cmu  coarse-fine  cocoa  code-dive  comics  commentary  communication  community  comparison  compensation  compilers  composition-decomposition  computation  computational-geometry  computer-memory  computer-vision  concept  conceptual-vocab  concurrency  config  confounding  confusion  consumerism  contrarianism  convexity-curvature  cooking  cool  correctness  correlation  cost-benefit  course  cracker-prog  critique  crosstab  crypto  cryptocurrency  curvature  d3  data  data-science  data-structures  database  dataviz  dbs  debt  debugging  decentralized  decision-making  decision-theory  deep-learning  definition  descriptive  design  desktop  devops  devtools  differential  differential-privacy  direct-indirect  dirty-hands  discipline  discovery  discussion  disease  distributed  diy  documentation  dotnet  duality  dumb-ML  dynamic  dynamical  early-modern  econometrics  economics  econotariat  ecosystem  editors  education  EGT  electromag  elegance  embodied  embodied-pack  engineering  ensembles  entanglement  entropy-like  erik-demaine  erlang  error  essay  estimate  europe  evan-miller  examples  exocortex  expert  expert-experience  explanation  exploratory  explore-exploit  exposition  facebook  faq  features  fermi  ffi  finance  fire  fisher  fitness  fluid  flux-stasis  focus  food  foreign-lang  fourier  frameworks  french  frequency  frequentist  frontend  functional  game-theory  games  gaussian-processes  gelman  gene-drift  generalization  generative  genetic-correlation  genetics  genomics  giants  git  gnon  golang  google  gotchas  government  grad-school  gradient-descent  graphical-models  graphics  graphs  ground-up  growth-econ  gtd  guide  GWAS  habit  hardware  hashing  haskell  health  hg  hi-order-bits  high-variance  higher-ed  history  hmm  hn  homo-hetero  howto  hsu  human-bean  hypothesis-testing  ide  ideas  identity  idk  IEEE  immune  info-dynamics  info-foraging  infographic  information-theory  inhibition  init  innovation  integration-extension  interface  internet  intricacy  intuition  investing  ios  iron-age  iteration-recursion  iterative-methods  jargon  javascript  jobs  jvm  kaggle  kernels  language  latent-variables  latex  latin-america  leadership  learning  learning-theory  lecture-notes  lectures  lens  lesswrong  let-me-see  levers  lexical  libraries  linear-algebra  linear-models  linearity  liner-notes  linguistics  links  lisp  list  llvm  lol  long-term  machine-learning  madisonian  magnitude  maker  management  maps  marginal  marginal-rev  markets  markov  math  math.CA  math.CO  math.DS  math.NT  measure  mechanics  medicine  medieval  mediterranean  memory-management  mental-math  meta:reading  meta:research  metabuch  methodology  metrics  michael-nielsen  microfoundations  migration  mihai  minimum-viable  mit  ML-MAP-E  mobile  model-class  models  moments  money  monte-carlo  multi  music  music-theory  negotiation  networking  neurons  nibble  nitty-gritty  nlp  noise-structure  nonlinearity  nonparametric  nostalgia  notetaking  numerics  nutrition  objektbuch  ocaml-sml  oceans  off-convex  oly  oly-programming  online-learning  oop  openai  operational  optics  optimization  ORFE  org:bleg  org:com  org:edu  org:fin  org:gov  org:junk  org:lite  org:mat  org:med  org:nat  organization  os  oss  osx  outdoors  overflow  p2p  p:null  p:someday  p:whenever  papers  parenting  paste  pdf  performance  personal-finance  phd  philosophy  phys-energy  physics  pic  plan9  planning  play  plots  pls  plt  poast  polynomials  pop-structure  popsci  population-genetics  pragmatic  pre-2013  prepping  preprint  presentation  princeton  prioritizing  priors-posteriors  privacy  problem-solving  procrastination  productivity  programming  progression  project  psychology  python  q-n-a  QTL  quantum  quantum-info  quixotic  quotes  r-lang  rand-approx  random  rat-pack  rationality  ratty  reading  realness  recommendations  reddit  redistribution  reference  reflection  regression  regularization  regularizer  reinforcement  replication  repo  research  research-program  retention  retrofit  rhetoric  right-wing  rust  s:*  s:***  saas  safety  sampling  sanjeev-arora  scala  scale  scaling-tech  scaling-up  scholar  scholar-pack  sci-comp  science  scitariat  search  sebastien-bubeck  securities  security  self-control  sequential  shipping  short-circuit  signal-noise  similarity  simulation  sinosphere  skeleton  skunkworks  sleuthin  slides  social  software  spanish  spatial  speaking  spotify  spring-2015  stackex  stamina  stanford  startups  state  statesmen  stats  stochastic-processes  stories  street-fighting  strings  stripe  students  study  studying  success  summary  summer-2014  survey  survival  sv  syntax  synthesis  system-design  systems  szabo  talks  taxes  tcs  tcstariat  teaching  tech  technical-writing  technology  techtariat  tensors  terminal  the-classics  the-great-west-whale  the-monster  the-trenches  things  thinking  tidbits  time  time-preference  time-use  tip-of-tongue  tools  top-n  track-record  tradeoffs  trees  trends  tribalism  tricki  tricks  trivia  truth  turing  tutorial  twitter  types  ui  unaffiliated  unit  unix  unsupervised  urbit  us-them  usa  ux  vague  values  variance-components  vcs  via:popular  video  visual-understanding  visualization  visuo  volo-avolo  waves  web  weightlifting  white-paper  whole-partial-many  wiki  wire-guided  wkfly  workflow  working-stiff  worrydream  worse-is-better/the-right-thing  writing  x-sports  yak-shaving  yoga  🌞  🎓  🔬  🖥  🦉 

Copy this bookmark: