Scripting Your Analysis

Loading...

Do curso por Johns Hopkins University

Reproducible Research

2998 classificações

Explore Related Videos

At Coursera, you will find the best lectures in the world. Here are some of our personalized recommendations for you

Johns Hopkins University

Reproducible Research

2998 classificações

Curso 5 de 10 em Specialization Data Science

This course focuses on the concepts and tools behind reporting modern data analyses in a reproducible manner. Reproducible research is the idea that data analyses, and more generally, scientific claims, are published with their data and software code so that others may verify the findings and build upon them. The need for reproducibility is increasing dramatically as data analyses become more complex, involving larger datasets and more sophisticated computations. Reproducibility allows for people to focus on the actual content of a data analysis, rather than on superficial details reported in a written summary. In addition, reproducibility makes an analysis more useful to others because the data and code that actually conducted the analysis are available. This course will focus on literate statistical analysis tools which allow one to publish data analyses in a single document that allows others to easily execute the same analysis to obtain the same results.

Na lição

Week 1: Concepts, Ideas, & Structure

This week will cover the basic ideas of reproducible research since they may be unfamiliar to some of you. We also cover structuring and organizing a data analysis to help make it more reproducible. I recommend that you watch the videos in the order that they are listed on the web page, but watching the videos out of order isn't going to ruin the story.

Introduction1:57

What is Reproducible Research About?8:19

Reproducible Research: Concepts and Ideas (part 1)7:11

Reproducible Research: Concepts and Ideas (part 2) 5:26

Reproducible Research: Concepts and Ideas (part 3) 3:26

Scripting Your Analysis 4:35

Structure of a Data Analysis (part 1)12:29

Structure of a Data Analysis (part 2)17:40

Organizing Your Analysis11:05

Conheça os instrutores

Roger D. Peng, PhD
Associate Professor, Biostatistics
Bloomberg School of Public Health
Jeff Leek, PhD
Associate Professor, Biostatistics
Bloomberg School of Public Health
Brian Caffo, PhD
Professor, Biostatistics
Bloomberg School of Public Health

[BLANK_AUDIO]

In this course, we're going to talk a lot about different

tools and techniques to help you make your work reproducible.

And now, before we get into all that, though, I just want to kind of

boil down the basic principle that we want to deal with throughout this whole course.

I mean, if there's one rule that I think we would, you would want to

adhere to, to make your work as reproducible

as possible is that you should script everything.

Scripting is really important and it's the basic idea of writing everything down.

Back in the old days you might have written down,

you know, things down on paper or in lab notebooks.

But now that we have more sophisticated computers and we can

program them, we write things down as pr, as scripts or programs.

So I mean, if you, as, just by way of analogy.

You think about you think about a great symphony or a great piece of music.

You think of if

you, if you imagine just hearing the melody of a piece of music, okay?

That's very interesting, it's also can be very beautiful.

But often, if you look at a symphony, there's a whole orchestra.

There's a whole set of musicians that are kind of backing up the melody.

And doing a lot of work and all that work is really important and contributes to

the texture, contributes to the, the kind of

feel of the whole piece of music, right?

So if you been out and if you were just hear

a symphony it probably would be difficult for you to reproduce that

sound even if you could hire your

own orchestra because you wouldn't know where all

the notes were, what all the, all the

different instruments were playing at any given time.

That even if you heard it a few times,

you saw, probably still wouldn't be able to do it.

And you, in particular you would definitely not be able to

reproduce that sound if you were to only hear the melody, right?

So, if you think about the presentation of of an

analysis or of your research [COUGH] the presentation is really

kind of like, like hearing the melody alone, right?

The melody is really nice, it's really important.

But it's, there's all this supporting material

that, that's going on behind the scenes.

All the, the analyses that you did, all the exploratory work,

all of the, kind of, dead ends that you went down, all the

decisions that you made in your analysis, which are not presented in

your, kind of, final presentation, whether it's a paper or a oral presentation.

But they're all very important.

So those are the supporting

players, all right?

So just like in music in research we need a way to

kind of write down everything that we did, everything that's going on right?

So in music this is typically called the score, right?

So the score is, is, is a way to kind of write

down in musical notation all the different things that are being played.

All the different things that are going on.

All the notes, all the rhythms.

All the timings, all the instruments, everything.

And so someone can look at the score and

say okay, well I need to put together all these things and

I, every- ever- people have to play at different times and different locations.

And so it's a way to, for you to

reproduce a piece of music as accurately as possible.

And, and musicians have kind of, have come, have

agreed upon a standard way to write down music.

And a standard notation for doing that.

And so this course is essentially about, you

know, defining and developing the notation of specifying

a re, a research project or a data analysis.

Right, so what is the, how do we write down what the what are the

things that we did, how did we talk about exactly what was analyzed, what was done.

And so the way that we do that essentially is through scripting, right?

We write down computer programs.

We talk, we we give instructions to the computer and tell it what

to do at any, at any given time with what type of data.

And so the main rule

for anything you in data analysis is to write a script.

And the more you can do that and the less you can kind of do things by, on

your own, by hand without writing it down, the

better and the more reproducible your work will be.

So this, so these computer programs, these scripts that we write.

These are like the score for your data analysis.

They tell someone else who's trying to reproduce

your work exactly what happened and what to do.

So, if you want to get a sense of what it means to script something.

You can just load

up Rstudio. And

when you load it up, you'll see that there's, that you, the

first thing you can do is you can open an R script.

So, here we do that and this is just a blank text file.

And then you can start writing R code in here.

And then you'll be writ-, you'll be programming, you'll be telling the

you know, R what to do in terms of a coding language, right?

So this is your, this, this is the, the easiest way to get started scripting.

Is to u, is to load up something like Rstudio and, and, and use the text

editor that's built in.

So, in this course we'll talk about a bunch of other tools things

like Nader, things like Markdown that'll help you kind of, organize your analyses.

But the bottom line is, to write it down, write

a script, write a program, and make your work reproducible.

Explore nosso catálogo

Registre-se gratuitamente e obtenha recomendações, atualizações e ofertas personalizadas.

O Coursera proporciona acesso universal à melhor educação do mundo fazendo parcerias com as melhores universidades e organizações para oferecer cursos on-line.

© 2019 Coursera Inc. Todos os direitos reservados.

Baixar na App Store

Baixar no Google Play