Data i guess

Exploring open available data   

[data-i-guess]

MOVED

I migrated all my data to this place instead :)

[data-i-guess]

Dying to win!

Every year we get the opportunity to vote on people who ”significantly improve the gene pool by eliminating themselves from the human race in an obviously stupid way” via the Darwin Award webpage. This has been an ongoing thing since 1993 and there are SO many nominees, like:

Rhino Poacher Killed By Elephant And Eaten By Lions (South Africa, 2019)

Masturbator Meets Hard End (Michigan, 2016)

Cigarette Lighter Triggers Fatal Explosion (Indiana, 1996)

I wanted to applaud the human effort by examining which countries has contributed the most in protecting the gene pool. 

Congratulations New Zealand, Croatia, Scotland, Australia and the U.S for top 5 placement in making our world a better place! 

Work flow

I collected the data from the Darwin Award’s webpage. All awards with an easy access of country origin was selected. In total, 59 countries have earned at least one Darwin Award the last 26 years (USA has info about each state but in this graph the data is treated as one country)

Separately, I collected population data (this data is static, but population has not spiked during this timeframe– so the image should be accurate)

I merged the two DataFrames into one and calculated the total Darwin awards per 1 million inhabitants. To avoid a skewed result, countries with a population smaller than 4 million was excluded. This made the total into 53 countries.

I wanted to visualize the data as a time series, therefore, I created a pivot table to be able to make a stacked bar chart. I then identified the top 15 countries, re-indexed the data, created the bar chart and added the background image!

[data-i-guess]

Often, we compete which country is the best (or worst) in something. To keep track of the score I hereby officially open 

THE BATTLE BETWEEN SPAIN & SWEDEN!

We will compete in very exciting subjects! Such as;

Parental leave (go Sweden!)

Life expectancy (I think Spain will win this one)

Gender wage gap

Education level

Cost of life

and other relevant and important topics!

The chance (risk?) of not dying a violent death is higher in Spain. Here, there were 0.66 homicides per 100,000 people in 2017. While in Sweden, there were 1.13 homicides per 100,000 people the same year. 

Point to Spain

Lets all also take a moment looking at Lithuania (4.53 homicides) and Latvia (4.15 homicides)...

Ps. Not all countries are represented in the data. This is related to either missing data or the focus of the graph. For full data source. 

[data-i-guess]

Makeover (isch) Monday!

A few week back the theme of Makeover Monday was top 10 chocolate bars. When scouting how to represent this the best I found someone who had made a bump chart over America’s kennel club most popular breeds - since I am a dog person I got curious: how does it look in Sweden?

Turns out: we all love the Labrador, but in Sweden we apparently also love hunting! (yes dog people, don’t be alarmed - I know the Labrador is part of hunting too, but they fetch). The largest fall from popularity is the Cavalier King Charles Spaniel, starting at rank 8 and rank 16 in 2018 - hm?! The most unexpected popularity journey is made by the Bichon Havanais (it’s a small fluffy dog)

It also saddens me that the breeds  with short noses still are so popular here, I hope a change in breeding will make Pugs, French bulldogs and other similar dogs able to breed. 

Work flow

I searched SKK for their yearly dog breed popularity data

gathered it, cleaned it and grouped it :)

then visualization time!

Ps. My dog isn't even on the list! How is that possible?? 

[data-i-guess]

MAKEOVER MONDAY

A few weeks back I participated in the online conference Demystifying Data Science 2019 and heard of MakeoverMonday.  

Needless to say, I took advantage of this amazing page to practice making my first R data visualisation!  

The data comes from this article 

Workflow

Checking and cleaning the data

Creating a function calculating frequency och organising data again

Trying to figure out R (Step 2 and 3 took a while.. ^^)

Trying out the visualisation on friends 

Improving the graph based on friends feedback

I am really happy with my first R try!

[data-i-guess]

160 years of Swedish feminist journal history

In Sweden, we have KvinnSam - National Resource Library for Gender Studies and university-wide research infrastructure.  KvinnSam’s main tasks are surveying and cataloguing literature on gender issues, and compiling and cataloguing manuscript material on women’s history, as well as providing reference services. Meaning, it hold  so much amazing data. 

Here, I have, with the help from KvinnSam, made a very simple visualisation of 160 years of Swedish literature history. 1880s til 1960 seems to be very productive time, with lots of different journals and many issues published/year. It is also humbling to see how Hertha, 150 years later, still has issues published today. 

Looking closer into the 1960 and onwards there are fewer journals and issues, but in an era of digitalisation it is quite remarkable how feminist journals continues and are created. 

I will look closer into the different subjects raised within the journals and how they are related to historical events.

You can find KvinnSam here

[data-i-guess]

Analyzing gender within STEM subjects has increased a lot; at least people talk a looot about it.

It made me curious on what has actually happen - what are the trends and how frequent are gender keywords in these studies for real?

I narrowed it down to check the two largest Engineering education journals I know (European Journal of Engineering Education and Journal of Engineering Education).

My search on these pages was:

Your search for [All: gender]

This gave 712 hits in total. However, it turns out that around only 100 articles had any keywords related to gender. Bummer..

To get some more data I added Engineering Studies which added around 60 hits and 20 articles having any related gender keywords.

I give you: 15 years of gender research within Engineering Education where the most frequent word (both over the year and counts) is GENDER.

It is interesting to see that underrepresentation and diversity are topics that have interested researchers during these 15 years. And also, that  feminist research, masculinity, intersectionality, trans and LGB have been examined since 2011.

[data-i-guess]

As promised, a visualization of human vs walkers’ killing count in Fear the walking dead.

The walkers start of kind of strong with almost 10 kills compare to the humans’ lame 5. But then they fall behind. So here you have it, the leading cause of death is due to humans within Fear the walking dead. 

The data held 111 deaths, 4 cause of death was missing - In total 107 deaths are visualized.

All in all, men beware of other men (dead or alive) if there ever is an undead apocalypse.

Data still from https://walkingdead.fandom.com/wiki/List_of_Deaths_(Fear)

[data-i-guess]

Since Fear the walking dead is more about sociology issues  than zombies I find it interesting.

I got curious who has, so far, survived the undead apocalypse most successful... turn out *drum-role* it is the cast read as women (I am leaving it unsaid if this is due to lack of diversity within the show or simply since the female characters follow reality and take less risk and, therefore, don’t die in stupid accidents ;)).

Fear the walking dead men!

Btw, it seems like the leading cause of death is caused by humans and not walkers.. I’ll do another visualization of this!

Data from: https://walkingdead.fandom.com/wiki/List_of_Deaths_(Fear)

[data-i-guess]

The stages of getting rejected for a PhD - a true story!

I found a tool online that I wanted to try (turned out not amazing - but hey!)

Phase 1 - Keeping it cool!

(I’m the cool individual somewhere in the middle, who was selected out of 100 people. I was very very cool. Indeed.)

Phase 2 - HUBRIS!!

(I might still be hard to spot, but I am the somewhat less cool individual who was selected)

Phase 3 - Reality check 

THE END

[data-i-guess]

I came across this article and wanted to visualize this passage the author is pressing on

“But laws are only part of the calculus in a woman’s ability to seek an abortion.”

Fun Alabama facts

There is roughly 5 million people living in the county

52% of the population are women

76% of women age 18-49 use any contraception method (pills, condoms, sterilization, withdrawal, rhythm, etc.)

Only one clinic in Alabama is open on weekends (I didn’t find any open on weekends, but google maps might be trolling me)

It might seem like there are some options, I mean 5 facilities on roughly 2.5 million individuals, which not all of them are in a reproductive age. However, according to the article:

“In 2014, 59% of women lived in Alabama counties without a clinic, compared with 39% in the rest of the United States, said Elizabeth Nash, senior state issues manager for the reproductive rights think-tank, Guttmacher Institute.”

Just read the article, it made me wanting to do a visualization to add to the author’s point of accessibility.

PS. Compare with for example Georgia (28),  California (522), NY (249) abortion facilities.

References;

https://censusreporter.org/profiles/04000US01-alabama/#fertility https://www.guttmacher.org/sites/default/files/report_pdf/state-level-estimates-contraceptive-use-in-us-2017.pdf

[data-i-guess]

So, I wanted to try to create a social network using Gephi. This is the result! I used data from https://listofdeaths.fandom.com/wiki/Game_of_Thrones  and I have here visualised who directly or indirectly killed someone (no, mass murder in cities is not counted...)