R & friends for better science in less time in big (& small) team collaborations

class: title-slide, right, bottom
background-image: url(img/horst-starwars-rey.png)
background-size: contain
background-position: bottom left

## R & friends for better science in less time in big (& small) team collaborations
### Julia Stewart Lowndes, PhD
### National Center for Ecological Analysis & Synthesis, UCSB

[ @juliesquid](https://twitter.com/juliesquid) 
[ lowndes@nceas.ucsb.edu](mailto:lowndes@nceas.ucsb.edu) 
[ openscapes.org/media](http://openscapes.org/media)

???

---
<img src="img/horst-lowndes-loop-wtext.png" width="100%" />

.footnote[
art: [@allison_horst](https://twitter.com/allison_horst)   
]

???

The way I have been thinking about it, open data science tools and practices and teamwork & collaboration exist in this feedback loop.

Learning and using similar software promotes and streamlines teammwork. And also working as a team better equips you to learn open practices data science.

But as env scientists, we're never taught data science/open practices, and we're not taught to work as a team, even when we are physically sitting with other scientists similar to us in our research groups research group

---
<img src="img/horst_openscapes_grassland_1500px-text-hex.png" width="100%" style="display: block; margin: auto;" />

.pull-left[
**A mentorship program for research teams to help build mindsets and habits around open data science, and grow the community of practice.** 
 
**Operated by NCEAS and incubated by Mozilla**

.footnote[
web: [openscapes.org](https://openscapes); twitter: [@openscapes](https://twitter.com/openscapes)
]

]

.pull-right[
<img src="img/Cohort1_zoom.png" width="85%" style="display: block; margin: auto;" />
]

???

this builds from nearly 7 years of the OHI

---
<img src="img/ohi-illustration-hex.png" width="80%" style="display: block; margin: auto;" />

.pull-left[
**A scientific method, tool, and community for channeling the best available scientific information into marine policy.** 
 
**Global assessments annually since 2012** 
 
**Independent assessments in ~20 geographies**

.footnote[
source: [*Halpern et al. 2012*: An Index to assess the health and benefits of the global ocean](http://www.nature.com/nature/journal/v488/n7413/full/nature11397.html); web: [ohi-science.org](http://ohi-science.org)
]
]

.pull-right[
<img src="img/ohi-team-2018b-crop.jpeg" width="100%" style="display: block; margin: auto;" />
]

???

A scientific method, tool, and community for channeling the best available scientific information into marine policy.

It is being used by the United Nations and by 20 groups around the world.

transition: But I like to talk about this with analogies from Star Wars

---

class: center, middle
background-image: url(img/horst-starwars-luke.png)
background-size: contain

.footnote[
art: [@allison_horst](https://twitter.com/allison_horst)
]

???

This is Luke Skywalker after he crashed his plane on Dagobah. He is there staring at a problem that he can’t solve with the skillsets he has, and it is demoralizing.

This plane can represent a lot of things for data analysis. 
As an individual, it can be data that is too big to be opened in Excel. 
It can be your self-taught code that you are too ashamed of embarrassed to share
And as a team, it can be not being able to run code sent by a colleague. 
It can be the time spent organizing the files your students send you through a mosaic of Google Drive and GitHub links, email attachments, and thumb drives.

For environmental scientists, there are ways big and small that can leave you feeling like Luke here.

If you imagine what it would be like for Luke if he had used whatever ropes or pulleys he had on hand, it probably wouldn’t have been pretty, reproducible, or gotten him where he needs to be on time.

transition: but luckily what happens next is that he meets Yoda

---
class: top, center
<img src="img/horst-starwars-yoda.png" width="88%" />

.footnote[
art: [@allison_horst](https://twitter.com/allison_horst)
]

???

But Yoda comes along and can solve Luke's problem in a way Luke never imagined. He uses the Force. And the Force is a powerful skillset and mindset because Luke can see this and learn from Yoda, and he can go on then to not only solve his current but it will broaden his mind to what is possible in the future.

Open data science is the Force that helps us do better science in less time. It can be:
As an individual: 
code in an open source language like R, having your own work be more automated and reproducible 
And as a team:

coding and coordinating collaboratively with tools like GitHub
being empowered by open practices and having it be part of the whole research process
These types of practices not only save time and are more reproducible, but it’s a whole mindset of what is possible as you reduce friction in the day-to-day analyses and become comfortable with the power of these tools.

So the Force is a powerful thing, but importantly, it is not something that exists on its own.

Transition: But Luke didn't go on to defeat the Empire himself, he had a whole community.

---

class: center, middle
background-image: url(img/horst-starwars-hands.png)  
background-size: contain

.footnote[
art: [@allison_horst](https://twitter.com/allison_horst)
]

???

And this community is powerful because of the diversity of backgrounds and expertise, and although not everyone is a Jedi, everyone contributes in really critical ways.

So in science this can be
Teaching scientists how to code through university courses and trainings like the Carpentries
Championing open science by mentioning it in science talks, publications faculty meetings, 
Funding opportunities for scientists to participate in open science
Building software to meet scientists where they are

transition: So to recap:

---

class: title-slide, right, bottom
background-image: url(img/horst-starwars-rey.png)
background-size: contain
background-position: bottom left

???

R is the Force that enables us as scientists to do better science in less time.

It empowers us to get our own data out of the swamp.

It empowers us and build off of our confidence and experiences and broaden the scope of scientific challenges that we can tackle, which for environmental scientists, includes food security, disease transmission, and climate change.

---
<img src="img/horst-lowndes-loop-wtext.png" width="100%" />

???

This is my outline. Will talk this through

---
<img src="img/OHI-hex.png" width="55%" style="display: block; margin: auto;" />

.footnote[
source: [Halpern et al. 2012](http://www.nature.com/nature/journal/v488/n7413/full/nature11397.html); web: [ohi-science.org](http://ohi-science.org); [@OHIscience](https://twitter.com/OHIscience)
]

The Ocean Health Index is a scientific endeavor to quantify impacts and benefits of oceans around the world using the best publically available data.

It is being used by the United Nations and by 20 groups around the world.

There is a lot to it but what is relevant here is that we combine lots of data, we repeat our analyses every year, and we do it as a team.

But we are marine scientists, and we were never trained to work responsibly with data.

So we found out the hard way that our default approaches were not reproducible by even ourselves.

Getting through this involved quite a reckoning, but when we got through it, we knew we had a story to tell.

---
.pull-left[
<img src="img/horst-starwars-luke.png" width="60%" style="display: block; margin: auto;" />

.footnote[
art: [@allison_horst](https://twitter.com/allison_horst), graphics: [Andy Teucher](https://github.com/ateucher/vic-geogeeks-2018-01-30)
]
]

.pull-right[

]

???

Because we were all self taught, we came at OHI with the mismatches. Difficult workflows, and difficult to collaborate.

---
class: middle, center
<img src="img/nature-screenshot-title.png" width="100%" style="display: block; margin: auto;" />

.footnote[
source: [Lowndes *et al.* 2017](https://www.nature.com/articles/s41559-017-0160); web: [ohi-science.org/betterscienceinlesstime](http://ohi-science.org/betterscienceinlesstime/)
]

???

Figuring out was a big enough deal that we published this in Nature Ecology & Evolution.

What made it unique: we shared our struggles, and described the transition

---
### More reproducibility & collaboration, less time
<img src="img/bsilt-fig1-simple.png" width="60%" style="display: block; margin: auto;" />

.footnote[
source: [*Lowndes et al. 2017* Our path  to better science in less time using open data science tools](https://www.nature.com/articles/s41559-017-0160)
]

???

It looks like this.

Shown here as the circles get smaller and moving up the y axis. It also go easier for us to collaborate each year, along the x axis.

We couldn't overhaul everything all at once, but focused on different pieces incrementally each year.

Having all of this coding infrastructure in place is what enables those 20 independent groups I mentioned to study ocean health in the places they care about.

---
class: center, middle
<img src="img/workflow-repro-wtext.png" width="100%" />