This was the result of the plant height being printed on the 3D printer. I experimented with whether having the plant height really tall and far apart, or shorter and more together would give better results. I feel the right one with the taller and more spread out data points turned out to be more successful in this situation.
For my second project where I created a lattice structure based off of the Hagberg Falling Number. What I did was use the Grasshopper definition used to calculate averages (created by Denton Fredrickson) and added on to it to make a definition used to create these graph structures for both location, and rye type. I then lined them up so that the graphs for location and rye type lined up and it created this interesting lattice structure which I then made into a physical object.
This is the grasshopper definition in Rhino which was used to create the graphs for the lattice structure. In the green area is where the agronomic data is pulled in and averages are generated (The Agronomic and Grain Quality Averages portion of these Grasshopper sketches were done by Denton Fredrickson, and I utilized the framework he had created to add on and create my projects). In the grey area to the right is the stuff I did to take those averages and create the agronomic data. I created points which were placed from this data, and then lines were drawn between the points to create the final result before lining them all up. This could have been more efficient but it worked for what I wanted to do and I was happy with it.
This is the result of a suggestion made by Denton to use ranges in my grasshopper definition. This led to some research on how the range node worked, and I was able to make this definition which was much smaller and more efficient than what I have shown before in the media samples. The nodes in the grey area were what I put in to make those graphs, and you can see that it is much smaller and to the point that what I was working with before.
The Long Drawing functions similarly to my first project, but instead graphs both ergot and protein percentage in response to a movement through decreasing Hagberg Falling Numbers. The cultivars and locations were again gridded, this time in order of average Hagberg Falling Number. The data is drawn with a point for each of the fourteen locations repeating through all fifteen cultivars (unlike in the painting, which is a movement through all 15 cultivars in each location sequentially). The song sounds with two overlaid melodies of 14 notes offset slightly, with the melody responding to protein % one octave higher than that of ergot %. Each note in the melody corresponds in frequency to relative protein/ergot percentage, within a specific location. The fourteen note order of locations stays the same while proceeding through each cultivar. All data is sounded relative to the whole set within two octaves of the B-flat minor key.
The painting contains a graph of % ergot ordered from average highest to lowest in both dimensions (cultivar vs. location), and a graph of % protein following the same order of cultivars and locations. The song which it accompanies is in the key of E minor, with fourteen 15-note melodies each divided by a full note rest. The variable of time is the movement through the 15 ‘cultivars’ (breeds) of rye from the least prone to ergot (fungal disease) to the most prone, repeating in structure through all of the studied locations from least prone to ergot to most (14 locations). The notes are determined by the relative percentage of protein divided through three octaves of E minor. Each registered breed is given a half note sustain, and each non-registered breed is given a quarter note. There is a bass drone under each 15-note melody which is the average for the current location across all breeds, dropped two octaves. It is meant to function generally as a graph of the correlation between ergot incidence (increasing as time) versus protein percentage (of grain).
This was my first attempt to visualize this new data set. I was initially interested in making some of the time-based data appear in a linear fashion, as this was something I hadn’t done yet in my data viz experiments. I decided to look at if there was a relationship between the date the heads emerged to its maturity date, and if that had an impact on the yield of those crops.
The overlapping circles are drawn for each rye type at each location, and the size of the circles are reflective of the yield. The four rows represent each province the rye was planted. From top to bottom: Alberta, Saskatchewan, Manitoba, and Ontario. I think it’s interesting how Alberta is much more consistent in maturity date and has higher yields than Saskatchewan or Manitoba.
This project was originally an experiment that revolved around showing data for every cultivar and every city at the same time. This, much like my first project, is written in Java on a program called “Processing”. This image created a sense of scale for a viewer of the size of rye kernels across the entire data set. Then using colour, I add another level of data by comparing instances of ergot growth across all of the cities and cultivars.
On the whole I was very impressed with this project. I was excited that it showed relationships between kernel size and location/cultivar that the scientist we had be working with confirmed are well known in rye breeding. I was extremely cool to learn that I was uncovering other, new relationships as well.
I would like to expand this project to cover more portions of the data. I was very content to have the size of the ellipses have a spatial connection with the size of the kernels. Upon more learning, I’ve decided that this visualization would be equally impactful when it contains other bits of data too.
Competition surrounds us. It is the driving force of the capitalist economy, it is often fuel for nationalistic, civic and institutional pride, and, from a Darwinian perspective, it is the reason for the existence of all living organisms. For Jamie Larsen, Crop Geneticist and Breeder with Agriculture and Agri-Food Canada, competition takes the form of expedited evolution through the selection and controlled breeding of wheat, rye, and triticale plants. In the case of his Fall Rye Co-operative study, when a particular fall rye variety performs well, such as high winter survival rates, high grain yields, and low amounts of ergot (a fungus toxic to humans), its seeds move forward to be used in the next year’s crop, which can then be bred with other top performers. By controlling the pollination of each crop of fall rye, Larsen is able to foster the development of a rye variety with increasingly ideal qualities over time. Competition at its finest!
Following this breeding process (which can take up to 12 years) the seed selection culminates most dramatically at the yearly Prairie Recommendation Committee meetings. Here, Larsen and his colleagues present the agronomic and grain quality data of their most promising seeds for evaluation, the goal being to have them ‘supported’ for registration with Canadian Food Inspection Agency, who gives the final approval for them to enter the marketplace. In the past, seeds would be supported or rejected based on a majority vote, but these days decisions are informed largely by the Agronomic Evaluation Team’s “Merit Assessment Tool” which scores each variety according to how a seed’s data measures up against other control varieties.
Once on the market, a fall rye variety will face a new set of competitive challenges from its new audience – the farmers of the Canadian Prairies. Even the most impressive seeds on paper must overcome the critical eye of the seasoned farmer and have enough appeal to inspire him or her to take a chance on the unfamiliar crop. Often, it is also a matter of convincing farmers of the merits of winter cropsor planting rye in general.
So, where does data visualization connect with the tumultuous world of competitive crop breeding? For myself, the Data Artist-in-Residence at the Disruptive Imaginings Data Visualization Lab at the University of Lethbridge, the connection materialized through a visual form familiar to modern sports enthusiasts – the bracket tournament.
Bracket tournaments are a pervasive type of competition in professional and league sports where all the winners of a set round of games move forward to the next round, and so on until the best team prevails through all rounds. During my research of bracket tournaments, I found a few delightful cross-overs with agriculture, most notably the reference to “planting” the “seeds,” which refers to the process of arranging the players/teams in the first set of brackets so that the strongest competitors do not meet until later in the tournament (1)(2). The practice of predicting the outcomes of bracket tournaments, known as bracketology, has become a sport in itself, taking over many offices and friendship networks during NCAA March Madness or NFL Fantasy Football. Since their domination of the sporting world, bracket tournaments have also become popular in many other domains, like movies, music, books, public radio, and now… fall rye!
The Fall Rye Leaguedata visualization will be shown on April 20th at 2PM at the University of Lethbridge. Incorporating interactive paint, a micro-controller, and lights, the multimedia project presents the statistical champion of Jamie Larsen’s Fall Rye 2013 – 2014 Cooperative study through a touchable cork interface. Location TBA.