Showing posts with label dataviz. Show all posts
Showing posts with label dataviz. Show all posts

Sunday, March 31, 2013

Picasso's Paintings


This post looks at the work that went into the Picasso graphic which won a gold medal at this year's Malofiej International Infographics awards. It was probably one of the most challenging and labour intensive pieces I've created. 

An exhibition of famed artist Pablo Picasso’s classic work came to  the Heritage Museum in Sha Tin, Hong Kong, last year. But the paintings and sculptures on display were just a tiny fraction of the work the Spaniard produced in his lifetime. The public may be aware of a few of his famous paintings which sold for millions at auction but we wanted to show his prolific career in more detail. 






Making the graphic

We decided to focus mainly on his paintings rather than go into detail on all of his other work. There were two reasons for this. First of all, the exhibition in Hong Kong was almost all paintings so it would be more relevant at the time. The second is that showing all types of work could overcomplicate the graphic, which was already becoming an ambitious piece to portray clearly. 

I wanted the main part of the graphic to work as a timeline showing all of his paintings and the styles he used. After some research we noticed the yearly painting count fluctuated quite a lot and so did the materials used to paint with. This is where I started. 

The source for this information was the Online Picasso Project, a comprehensive database of all documented Picasso artwork and artefacts by Dr. Enrique Mallen, a professor at Sam Houston University in the U.S. Every painting had a detailed profile including materials used, dates, dimensions and so on. After contacting Dr. Mallen I was allowed access to the data which we then processed and compiled into a spreadsheet, which took quite a while. I had some help from our graphics coordinator/researcher on this too. From there I laid out the data in a few different types of chart, experimented but decided to go with a representation of each individual painting as a dot, stacked by year.

The main reason for this is that I wanted to show information on each painting individually. We had access to this very detailed information so it would be a waste just to combine all of these paintings into a yearly total, a bar chart for example. Having each as a dot also allowed me to keep them in chronological order. This would help show any patterns or stories that may emerge as I worked further into the graphic.



First of many spreadsheets. This one for totals - dots




Counting dots



One piece of information we could include on each painting was the material used to paint it. So I went through and split the dots into two colours depending on whether they were mostly oil, watercolour or gouache. Some were a combination so in this situation I used the medium which was listed as being used most or most dominant. Any pieces which used neither such as purely pen, pencil, pastel etc fell into a different category so weren't counted as paintings.

Another layer of information we wanted to include was prices which his paintings sold for. The source for this would be ArtPrice, an online subscription company which compiles and updates art reference databases that cover art auction prices and images from its library of 290,000 auction catalogs.

We had to go through every Picasso painting they had on record and match it to the paintings listed in order on the chart and spreadsheet. Again, this was a huge amount of work as they all had to be sifted through manually using a combination of photos of the painting, dates, title of the piece and sometimes things like exact dimensions and registered owners. Some paintings had the same title and looked similar so they needed extra checking to be absolutely sure we had the correct match.

After all of this was compiled I experimented with ways to show the auction price data, including different weight/shades of the dots, but decided softer separate circles was the way to go. They were drawn up one year at a time then placed on the correct dots.



One year's circles drawn up



I decided on circles because it is easier at first glance to pick out big sales and also dense areas of sales where circles are close to each other or overlap, suggesting popular periods of his work at auction. 

Other information we wanted to include was which paintings were actually on show at the museum in Hong Kong. This would just be a secondary layer and not shown in too much detail so not to clutter the graphic. The names of those paintings were left off but highlighted using a thin black line around some of the dots. This gives a sense of when those particular pieces were painted and oil or watercolour/gouache. It is just as important to show restraint and know what to leave out of a graphic as it is to know what to include.

Picasso's paintings only account for around a quarter of his lifetime's work. A breakdown of his other works such as sculptures, ceramics drawings etc. was addressed with a simple but clear bar chart near the top of the graphic.

Of course all of these elements didn't just slot into place nicely. It takes hours, sometimes days of playing around with them. Trying things in different locations, order and size. Specially the key. This is one of the most important parts of the graphic. If the key isn't clear to the reader then they won't understand the rest of the graphic. I also think it's good to put a few words explaining the concept behind the graphic and how to read it.





Under the timeline we wanted to show more general periods of his work rather than individual pieces. Picasso's style of painting and use of colour changed a lot over his long career. We thought it was important to explain this thoroughly and also show some examples of his paintings. This would also help the reader connect the dots to real artwork. Rather than looking at only the data. A lot of attention was given to the annotation here with plenty of edits. We show information visually but when text is included it's just as important to try and perfect.





 The graphic ran as a full broadsheet back page during the exhibition.





Saturday, February 16, 2013

Asteroids


A 45-metre-wide asteroid came remarkably close to Earth on Friday, even closer than communication and weather satellites. It was be the nearest known close miss for an object of its size.   
When this story was first mentioned in the newsroom, a few days before the incident, it sparked debate. People were intrigued as to how close these objects come to Earth. How many pass by? And how fast or large are they? A perfect opportunity for an interesting graphic.    
As usual, NASA had every piece of information we needed. Their Near-Earth Object Program was established in 1998 to help coordinate, and provide a focal point for the study of comets and asteroids that can approach the Earth's orbit. They have data sets on all close approaches to Earth since 1900 and projected forward to 2200.    




The main part of the graphic shows all close approaches passing the Earth at a distance of one Lunar Distance (LD) or less. In other words, passing closer to Earth than the Moon. All 199 historical and projected passes are shown. All are arranged on the vertical axis by the distance they came to Earth. The axis represents the distance from the Earth to the Moon. Both of which are represented at each end, drawn to scale. The Length of each bar represents the speed at which the asteroid was traveling. White objects have already passed and orange are forecast.     





We also included two smaller diagrams. One showing 2012 DA14's orbit and how it will pass Earth and another showing its size compared to the Space Shuttle and the largest asteroid on the chart.    





The chart below did not make it on to the graphic.    





This shows all the asteroids over time. Every close approach recorded by NASA from the year 1900 to 2200 going out even further to 5 Lunar Distances or less. When we plotted it on this chart we noticed a strong trend. The last decade or so has seen a huge spike in the number of close approaches. Or has it? We figured this chart was too good to be true and assumed it may have something to do with recent technology and a greater ability to track these objects now. After speaking to NASA our suspicions were confirmed. It is harder to back track and accurately plot every close approach earlier in the 20th century and hard to predict as many in the distant future. After learning this we decided the chart was slightly misleading and decided to drop it.    
We decided the information we were showing was strong and simple enough to hold a full page and ran the graphic in Friday's newspaper as a back page.





Wednesday, January 30, 2013

Arteries of the city


I created this graphic back in October shortly after Hong Kong's Transport Department released their annual traffic census. The 182-page report was packed with statistics and useful information but the best data was a thorough breakdown of daily traffic on almost every stretch of road. The territory has hundreds of counting stations. We thought the chance to do something with these numbers was too good to miss so we pitched an analysis to the Editor who was keen to give it a good space.








The thickness of each line represents the average daily number of vehicles traveling on that stretch of road. The colour represents the percentage change on last year. This helps the reader to explore the data in two ways. You can clearly see that the main roads along the front of Hong Kong Island are still the busiest. But by the colour, the traffic has also mostly decreased compared to last year. You can also see that a lot of the roads in Kowloon have become busier. In particular the area up to the left, Tsuen Wan, the three tunnels north and the major highways to the east.

There is also a clear change in the cross-harbour tunnels. The Eastern and Western Harbour Tunnels have increased in traffic compared to the central tunnel which has decreased. But they still see less traffic.

The graphic is also a fun way for the reader to take a look around their neighbourhood or route to work.



Friday, January 18, 2013


Roadside pollution in Hong Kong hit a record high this year, on August 2 in Central, with the index reaching 212. As the year drew to a close, we took a look at how the air had fared in every hour in the year up to the publishing date.




Each day is represented by a row of 24 squares, one for each hour of the day. The shade of the square indicates the pollution reading at Central monitoring station.




The idea was to give the reader an overall glance at the year so they can see the worst and clearer periods, but also the opportunity to dig a little deeper into the hourly data. We also added some text pointers to explain some probable contributing factors to clearer air such as typhoons.









Hong Kong’s air pollution is often blamed on its proximity to mainland China’s industry. Wind direction each day is included as small arrows to the left of each row in order to gauge, if any, the relationship between northerly winds and bad air.





We wanted the reader to come to their own conclusion about the relationship with wind direction but it turns out a lot of the bad spells coincide with winds from the North (black arrows). See below.





The graphic was published as a full broadsheet back page near the end of December.





Thursday, December 13, 2012

Messi v Muller

We published this information graphic when Lionel Messi broke Gerhard Muller's record of 85 goals in a calendar year. 





The chart shows every goal, plotted by date scored. The length of a bar represents the number of goals scored in that day’s match. 





We wanted to do more than simply compare the number of goals each player has scored. Presenting every goal on a timeline in this way allows us to study the data and draw our own conclusions as to which of the two goalscorer's records in most impressive. It lets us see busy scoring periods, drier seasons, big hauls in some games (often scoring 4 or 5 goals) and most importantly, how Muller continued his prolific scoring rate for many years after he set the record in 1972.






The data for Messi's goals and when he scored them was pretty easy to find from various reliable websites. Muller's was not so easy. So much so we had to commission a private sports data company in Germany to compile it for us. In the end we decided it was worth the extra expense in order to show the information this way and give the reader a different angle to other Messi infographics they may have seen recently.

Sunday, December 2, 2012

As the government considers a revamp of electricity tariffs, we took a look at Hong Kong's power consumption. Who uses all the electricity in our city and what is it used on? This graphic was printed last week as a full broadsheet back page. Data set was provided by the government's Electrical and Mechanical Services Department.






This type of chart is known as a Sankey diagram. The thickness of each line reflects a value. In this case, an amount of electricity in terajoules. All of the lines add up to give subtotals and totals by users (grey) and end use (coloured).




Friday, November 23, 2012

Wages in Hong Kong

As Hong Kong’s wealth gap widens, who is on which side of the growing divide? This graphic shows which industries pay the most and which have the biggest discrepancies between their highest and lowest earners. The dataset was provided by the Hong Kong Census and Statistics Department and covers all official employment in HK. This graphic was printed as a full broadsheet page.







Each industry is represented by a series of 3 connected dots. The middle dot is the median wage for that sector and the top and bottom dots are the averages for high and low earners (90th and 10th percentile). This shows some industries have a huge difference between top earners and what most others are paid.






The height of the dots reflects the hourly wage in HK$. The industry, or green line/dots, is also placed left to right along the x axis according to how many people work in that industry. You can see Import and Export Trade alone out to the right of the chart. By far the biggest employment sector in Hong Kong.







Restaurants and estate services had the lowest two averages.