15 Principles for Data Scientists

marksalen's avatarOpen Source Research

I have developed 15 principles for my daily work as a data scientist. These are the principles  that I personally follow :

1- Do not lie with data and do not bullshit: Be honest and frank about empirical evidences. And most importantly do not lie to yourself with data

2- Build everlasting tools and share them with others: Spend a portion of your daily work building tools that makes someone’s life easier. We are freaking humans, we are supposed to be tool builders!

3- Educate yourself continuously: you are a scientist for Bhudda’s sake. Read hardcore math and stats from graduate level textbooks. Never settle down for shitty explanations of a method that you receive from a coworker in the hallway. Learn fundamentals and you can do magic. Read recent papers, go to conferences, publish, and review papers. There is no shortcut for this.

4- Sharpen your skills: learn one language well…

View original post 413 more words

F# Weekly #21 2013

Welcome to F# Weekly,

A roundup of F# content from this past week:

News

Videos/Presentations

Blogs

That’s all for now.  Have a great week.

Previous F# Weekly edition – #20

F# Weekly #20 2013

the lhs of an F# developer’s desk

the rhs of an F# developer’s desk

Welcome to F# Weekly,

This past week was full of interesting events and happenings. If you follow F#, you must read all of this carefully. The roundup of F# content:

News

Videos/Presentations

Blogs

That’s all for now.  Have a great week.

Previous F# Weekly edition – #19
lamdbaLadies

Three easy ways to create simple Web Server with F#

I have tried to find easiest ways to create a simple web server with F#. There are three most simple ways to do it.

The goal is to create a simple web service that maps web request urls to the files in the site folder. If file with such name exists then return its content as html. Assume that all html files located in ‘D:\mySite\‘.

HttpListener

First and probably the most promising option was created by Julian Kay and described in his post “Creating a simple HTTP Server with F#“. I slightly modified source code to satisfy my initial goal. You can find detailed description of how it works in Julian’s post. (Works from FSI)

open System
open System.Net
open System.Text
open System.IO

let siteRoot = @"D:\mySite\"
let host = "http://localhost:8080/"

let listener (handler:(HttpListenerRequest->HttpListenerResponse->Async<unit>)) =
    let hl = new HttpListener()
    hl.Prefixes.Add host
    hl.Start()
    let task = Async.FromBeginEnd(hl.BeginGetContext, hl.EndGetContext)
    async {
        while true do
            let! context = task
            Async.Start(handler context.Request context.Response)
    } |> Async.Start

let output (req:HttpListenerRequest) =
    let file = Path.Combine(siteRoot,
                            Uri(host).MakeRelativeUri(req.Url).OriginalString)
    printfn "Requested : '%s'" file
    if (File.Exists file)
        then File.ReadAllText(file)
        else "File does not exist!"

listener (fun req resp ->
    async {
        let txt = Encoding.ASCII.GetBytes(output req)
        resp.ContentType <- "text/html"
        resp.OutputStream.Write(txt, 0, txt.Length)
        resp.OutputStream.Close()
    })
// TODO: add your code here

Self-hosted WCF service

The second option is a tuned self-hosted WCF service. This approach was proposed by  Brian McNamara as an answer to the StackOverflow question “F# web server library“. (Works from FSI)

#r "System.ServiceModel.dll"
#r "System.ServiceModel.Web.dll"

open System
open System.IO

open System.ServiceModel
open System.ServiceModel.Web

let siteRoot = @"D:\mySite\"

[<ServiceContract>]
type MyContract() =
    [<OperationContract>]
    [<WebGet(UriTemplate="{file}")>]
    member this.Get(file:string) : Stream =
        printfn "Requested : '%s'" file
        WebOperationContext.Current.OutgoingResponse.ContentType <- "text/html"
        let bytes = File.ReadAllBytes(Path.Combine(siteRoot, file))
        upcast new MemoryStream(bytes)

let startAt address =
    let host = new WebServiceHost(typeof<MyContract>, new Uri(address))
    host.AddServiceEndpoint(typeof<MyContract>, new WebHttpBinding(), "")
      |> ignore
    host.Open()
    host

let server = startAt "http://localhost:8080/"
// TODO: add your code here
server.Close()

NancyFx

The third one is based on NancyFx. It is lightweight, low-ceremony, framework for building HTTP based services on .Net and Mono. Nancy is a popular framework in C# world, but does not have a natural support of F#. The F# code looks not so easy and simple as it could be. If you want to make it work, you need to create console application and install the Nancy and Nancy.Hosting.Self NuGet packages.

module WebServers

open System
open System.IO
open Nancy
open Nancy.Hosting.Self
open Nancy.Conventions

let (?) (this : obj) (prop : string) : obj =
    (this :?> DynamicDictionary).[prop]

let siteRoot = @"d:\mySite\"

type WebServerModule() as this =
    inherit NancyModule()
    do this.Get.["{file}"] <-
         fun parameters ->
              new Nancy.Responses.HtmlResponse(
                  HttpStatusCode.OK,
                  (fun (s:Stream) ->
                      let file = (parameters?file).ToString()
                      printfn "Requested : '%s'" file
                      let bytes = File.ReadAllBytes(Path.Combine(siteRoot, file))
                      s.Write(bytes,0,bytes.Length)
              )) |> box

let startAt host =
    let nancyHost = new NancyHost(new Uri(host))
    nancyHost.Start()
    nancyHost

let server = startAt "http://localhost:8080/"
printfn "Press [Enter] to exit."
Console.ReadKey() |> ignore
server.Stop()

Further reading

Live tweets from Alea.cuBase F#-for-financial-GPU event in London.

Yesterday, there was Alea.cuBase live coding session. Don Syme led live tweets from inside. This post is dedicated to those who missed it.

F# Weekly #19, 2013

Welcome to F# Weekly,

Tiny weekly this time, a short roundup of F# content from this past week:

News

Videos/Presentations

Blogs

P.S. The top 20 data visualisation tools

That’s all for now.  Have a great week.

Previous F# Weekly edition – #18