Spotify: https://open.spotify.com/user/1235450353/playlist/4jqNCyUylGh9zCYt0JHqEv...
YouTube: https://www.youtube.com/playlist?list=PLSUh6oJ5ZotVoTIXr6OBeLuUSor-PDN_p

• The Seed - ANIMA!
• Looped - Kiasmos
• Naive - RKCB
• Making Distractions - Kllo
• Lights - YOSIE, LO
• CR-V - Cuco
• Arisen My Senses - Björk
• Hush - Tourist
• You and Me - Shallou
• Last Day… - Shallou
• Life, Life (Andy Stott Remodel) - Ryuichi Sakamoto
• Nix - Dawn of Midi
• A Meaningful Moment Through a Meaning(less) Process - Stars Of The Lid
• Obfuse - Floating Points

I wanted give Expecty a try, so I forked the repo to eed3si9n/expecty, added sbt build, patched up the code so it works with 2.10, 2.11, 2.12, and 2.13.0-M4, sent a few pull requests upstream, changed the package name, and published my fork to Maven Central:

libraryDependencies += "com.eed3si9n.expecty" %% "expecty" % "0.11.0" % Test

and for Scala.JS and Scala Native:

libraryDependencies += "com.eed3si9n.expecty" %%% "expecty" % "0.11.0" % Test

Here's how we can use this:

scala> import com.eed3si9n.expecty.Expecty.assert
import com.eed3si9n.expecty.Expecty.assert
 
scala> assert(a * b == 7)
java.lang.AssertionError:
 
assert(a * b == 7)
       | | | |
       1 3 3 false
 
  at com.eed3si9n.expecty.Expecty$ExpectyListener.expressionRecorded(Expecty.scala:25)
  at com.eed3si9n.expecty.RecorderRuntime.recordExpression(RecorderRuntime.scala:34)
  ... 38 elide

As you can see, you get a nicer error message automatically.

Source dependencies is one of features that existed in sbt since ever, but hasn't been documented well.

Using sbt-sriracha, we can define a hybrid dependency to µTest as follows:

lazy val utestJVMRef = ProjectRef(uri("git://github.com/eed3si9n/utest.git#5b19f47c"), "utestJVM")
lazy val utestJVMLib = "com.lihaoyi" %% "utest" % "0.6.4"
 
lazy val root = (project in file("."))
  .sourceDependency(utestJVMRef, utestJVMLib)
  .settings(
    name := "Hello world",
    testFrameworks += new TestFramework("utest.runner.Framework"),
  )

Now using these changes I can run custom µTest on Scala 2.13.0-M4.

Yesterday I wrote about cross JVM testing using Travis CI.

testing Scala apps on macOS using Travis CI

Here's how we can test Scala apps on macOS using Travis CI. This is adapted from Lars and Muuki's method: Testing Scala programs with Travis CI on OS X

dist: trusty
 
language: scala
 
matrix:
  include:
    ## build using JDK 8, test using JDK 8
    - script:

Oracle is moving to ship non-LTS JDK every 6 months, and LTS JDK every 3 years. Also it's converging to OpenJDK. In this scheme, JDK 9 will be EOL in March 2018; JDK 10 will come out in March 2018, and EOL in September 2018; and LTS JDK 11 that replaces JDK 8 in September 2018 will stay with us until 2021.

As we will see quick succession of JDKs in the upcoming months, here's a how-to on testing your app on JDK 8, JDK 9, and JDK 10 Early Access using Travis CI.

dist: trusty
 
language: scala
 
matrix:
  include:

This post introduces sbt-nocomma that introduces nocomma { ... } macro that removes commas from build.sbt:

import Dependencies._
 
lazy val root = (project in file("."))
  .settings(nocomma {
    ThisBuild / organization := "com.example"
    ThisBuild / scalaVersion := "2.12.4"
    ThisBuild / version      := "0.1.0-SNAPSHOT"
 
    name := "Hello"
 
    // comment works
    libraryDependencies += scalaTest % Test
  })

As tweeted, I made some sbt stickers.

I made these for myself, but I've been sharing them with the contributors to the sbt ecosystem.

If you've ever reported a bug, answered Stackoverflow question, helped someone else's build problem, created an sbt plugin, given a talk about sbt, or sent in a pull request, please come say hello at ScalaMatsuri or Scala Days.

Wrote herding cats: day 17 featuring initial and terminal objects, product, duality, and coproduct.

Model representation

x⁽ⁱ⁾ denotes input variables (feature vector), and y⁽ⁱ⁾ denotes the output (target variable). A pair (x⁽ⁱ⁾, y⁽ⁱ⁾) is called a training example. A list of 1 .. m training example is called a training set.

Supervised learning can be modeled as a process of coming up with a good function h called hypothesis that predicts y from an unknown x.

For instance, given square footage, predict price of a house in some city. In the above, h_θ(x) is modeled as a linear function. The hypothesis h_θ(x) = θ₀ + θ₁ x is called linear regression with one variable (or univariate linear regression).

This is part 3 of sbt server integration series. I've covered VS Code and Sublime Text 3 thus far.