diff --git a/img/dvis.jpg b/img/dvis.jpg new file mode 100644 index 0000000..6bc82cf Binary files /dev/null and b/img/dvis.jpg differ diff --git a/img/ps.png b/img/ps.png new file mode 100644 index 0000000..f26abf8 Binary files /dev/null and b/img/ps.png differ diff --git a/img/vislapse.gif b/img/vislapse.gif new file mode 100644 index 0000000..2ecd8f1 Binary files /dev/null and b/img/vislapse.gif differ diff --git a/img/week10.png b/img/week10.png new file mode 100644 index 0000000..39f6a41 Binary files /dev/null and b/img/week10.png differ diff --git a/img/week11.png b/img/week11.png new file mode 100644 index 0000000..6dcbcb4 Binary files /dev/null and b/img/week11.png differ diff --git a/img/week12.png b/img/week12.png new file mode 100644 index 0000000..5634f5a Binary files /dev/null and b/img/week12.png differ diff --git a/img/week13.png b/img/week13.png new file mode 100644 index 0000000..6dd093b Binary files /dev/null and b/img/week13.png differ diff --git a/img/week4.gif b/img/week4.gif new file mode 100644 index 0000000..595cd87 Binary files /dev/null and b/img/week4.gif differ diff --git a/img/week5.png b/img/week5.png new file mode 100644 index 0000000..12d4f3b Binary files /dev/null and b/img/week5.png differ diff --git a/img/week6.jpeg b/img/week6.jpeg new file mode 100644 index 0000000..6e0bce1 Binary files /dev/null and b/img/week6.jpeg differ diff --git a/img/week8.png b/img/week8.png new file mode 100644 index 0000000..b79f73e Binary files /dev/null and b/img/week8.png differ diff --git a/img/week9.png b/img/week9.png new file mode 100644 index 0000000..97ba10c Binary files /dev/null and b/img/week9.png differ diff --git a/week1.md b/week1.md index e69de29..b7b567f 100644 --- a/week1.md +++ b/week1.md @@ -0,0 +1,19 @@ +# A Wordcloud of Every Inaugural Address in the History of the United States (1789-2021) + +![image](https://github.com/mike3osei/reflections-1/blob/master/img/dvis.jpg) + +I can honestly say that up until this day, I had never intentionally looked up an actual data visualization in my life before. Coming from a CS, UI/UX background that can probably come off as a shock to most. One thing I can say is that I do take visual design very seriously so it was definitely hard to pick one visualization. Something that did stop me while scrolling and looking through whats out there was this visualization of United States Inaguaral Addresses from 1789 to 2021. + +## Link +https://www.reddit.com/r/dataisbeautiful/comments/l79w12/oc_a_wordcloud_of_every_inaugural_address_in_the/ + + +## Initial Response + +Initially I did think it was pretty powerful how they emphasized certain words throughout certain addresses. It did make me pretty skeptical that there was definitely a big bias involved because the words that were emphasized seemed to be from the designers point of view versus the speakers point of view. I could be wrong about that but that was my initial assumption. + +I did find it super impressive that someone like myself, who probably won't be reading any of these addresses anytime soon, was able to take something away from that specific time period. That definitely was powerful and pretty big for me. + +## Takeaways + +I immediately took away something from my experience interacting with this visualization which is something I defintiely want to do with my own future visualizations. I think thats definitely major as a designer, being able to have such impact. diff --git a/week10.md b/week10.md new file mode 100644 index 0000000..4358335 --- /dev/null +++ b/week10.md @@ -0,0 +1,16 @@ +# Research and Application of Dynamic and Interactive Data Visualization Based On D3 + +![image](https://github.com/mike3osei/reflections-1/blob/master/img/week10.png) + +This publication describes the entire procedure of dynamic and interactive data visualization from several angles. They use D3 as a basis of this research and look for different advantages or techniques that can make visualization interactive data better. This analysis is conducted by the School of Information Technology, Shanghai Jianqiao University. + +## Link +https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=7846608 + +## Initial Response + +Intitially I found it very intersting how they used different mediums to go about analyzing data-driven visualizations. They mentioned the use of SVG and compared its strong-suites to D3. I personally never heard of anything called a D3 Visualization model but it was interesting hearing how they broke up the workflow into, requirement analysis, data analysis, visualization, and web visualization. They broke up alot of the specific nodes involved in application models into data application layers that reflect back on the actual topic itself. It was interesting seeing their bar chart examples as they were flat but still multidimensional. I was personally pretty interested about how they managed to implement this model. + +## Takeaways + +I thought it was interesting how they chose to represent their vis' as either grouped or stacked in certain examples. The key technolgoy for interactive or dynamic visualizations is very important in data-driven visualization. I wouldn't think to immediately think about that realm, but it was important to bring up for sure. \ No newline at end of file diff --git a/week11.md b/week11.md new file mode 100644 index 0000000..5bd87a9 --- /dev/null +++ b/week11.md @@ -0,0 +1,16 @@ +# DATAVIS HELPER: A TOOL FOR EXPLORING THE DESIGN SPACE OF DATA VISUALIZATION + +![image](https://github.com/mike3osei/reflections-1/blob/master/img/week11.png) + +This publication describes the development of a tool that could be used in academia as well as the real world. This tool was meant for exploring the design space of data visualization. They state that the idea is to develop a web tool that applies the principles of visual encoding. The testing was done in two cohorts. They were looking for people without any specific datavis background. + +## Link +http://purl.utwente.nl/essays/73188 + +## Initial Response + +Intitially I found it very intersting how they managed to developed a platform meant for developing these forms of visualization. The UI was typical with graidents and so forth, but I was very unimpressed by how the userflow worked. I think with datavis it's important to be able to guide users. + +## Takeaways + +I thought it was interesting how they managed to develop their inferface design for their testing group. When doing visual analysis its important to have a great UI that is simple to use. A lot of the target audience here was initial users that may not be comfortable with an environment like virtual testing. It's important to note that I feel that this research paper didn't do a great job in that regard and I wonder if that would affect the end results. \ No newline at end of file diff --git a/week12.md b/week12.md new file mode 100644 index 0000000..5f7352e --- /dev/null +++ b/week12.md @@ -0,0 +1,15 @@ +# Visualization of Graphs with Associated Timeseries Data + +![image](https://github.com/mike3osei/reflections-1/blob/master/img/week12.png) + +The publication describes that overlaying data on graph vertices, one value at a time can lead to faster performance for tasks involving analysis of a graph that involves time series data. By looking at the vertices and deciding whether to use multiple, or singular datapoints there comes an effect on their said oucome. Its usually an effect that promotes more clarity in the graphs visualizaiton according to this group. + +## Link +https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=1532151 + +## Initial Response +I thought it was interesting that the number of nodes at a vertex could directly effect the ability for an individual to perceive data. To me, I wonder if there is a difference between overwhelming the user and visually adding more actions to the vertecies. Overlaying, data on a graph in this fashion is interesting because it does appear that it would not be as efficeint in visual search. I would think that this would be visually distracting, but in many cases, according to this study it was faster and more accurate. + +## Takeaways + +I think my biggest takeaway is the impact of user testing in this study. I think my initial thoughts are to not judge a book by its cover, but in this case there was insights that were presented to bring forth the not expected outcome. Single vs. Multiple attributes also seemed to have different levels of accuray but the more nodes represented the easier it was for the eyes. diff --git a/week13.md b/week13.md new file mode 100644 index 0000000..23fa404 --- /dev/null +++ b/week13.md @@ -0,0 +1,14 @@ +# Graphical representation of multivariate data using Chernoff faces + +![image](https://github.com/mike3osei/reflections-1/blob/master/img/week13.png) + +The article describes the use of cartoon-like faces to represent value changes in a dataset. According to this article, they implemented a Stata implementation of a face-generating algorithm. Apparently Chernoff proposed a way of showcasing multivariable data as cartoon faces. They describe this as a way of generating results that are more simple to remember. + +## Link +https://journals.sagepub.com/doi/pdf/10.1177/1536867X0900900302 + +## Initial Response +I thought it was interesting the use of scale in the context of data visualization. I didn't think that a face, given its emotions, complexity, and features could give out that many expressions. I found it interesting that something as simple as outliers can be hard to measure due to the cernoff measurements. + +## Takeaways +I found it interesting that they were pretty optimistic about the future interations of the Chernoff face. They mentioned how it would evovle if they had had time to fine-tune it. For instance, having a limits implementation would be very important in said case. Faces, can help detect trends and patterns which can be very useful moving forward. diff --git a/week2.md b/week2.md index e69de29..5f2cff0 100644 --- a/week2.md +++ b/week2.md @@ -0,0 +1,16 @@ +# Average Police Salary by State | Accessibility + +![image](https://github.com/mike3osei/reflections-1/blob/master/img/ps.png) + +One thing that I have noticed that can be very important in visual design is the use of color. This may seem obvious, but in most cases people use color for purely aesthetic or hierarchical representations. Being someone who is color blind, the use of color in most graphs is pretty trash, in my opionion. On a decent amount of visualizations, I never seem to truly understand why the colors were of any significance. In this example, the use of color is definitely very important and accessibility is accounted for in a very wise manner. + +## Link +https://www.reddit.com/r/dataisbeautiful/comments/lg2mo8/oc_average_police_salary_by_state/ + +## Initial Response + +Initially when seeing this visualization, my first thought that came to mind, was that the use of gradients was actually very effective in pointing my eyes in the right direction. Especially when seeking a region with higher pay versus one with lower pay. Another observation I definitely made was that places with big cities were probably more in tune to have an average police salary that was on the higher side. One great thing about this visualization that I noticed was the positioning of state nodes. They were pretty much in the same position they usually would be. That made it very easy to traverse the items. I already had a key-value map of the United States stored subconciously. One that I was already familiar with. + +## Takeaways + +Is there much data visualization can do in terms of accessibility? I feel like in computer science, with web development and mobile development, accessibility is a big part of the workflow. It is very important to make good use of font sizes, colors, and tools that can be widely adopted across your audience. diff --git a/week3.md b/week3.md index e69de29..a42a665 100644 --- a/week3.md +++ b/week3.md @@ -0,0 +1,16 @@ +# One Year Animation of the 7-day average new Covid-19 cases per capita (NYT County dataset) + +![image](https://github.com/mike3osei/reflections-1/blob/master/img/vislapse.gif) + +The Covid anniversary is approaching us, and we're around that time where we can look at some of these interesting visualizations that were developed over the past year or so. + +## Link +https://www.reddit.com/r/dataisbeautiful/comments/lcn11g/oc_1_year_animation_of_the_7day_average_new/?utm_source=share&utm_medium=web2x&context=3 + +## Initial Response + +Initially when seeing this visualization, the first thought that came to my mind, was that timelapses really can tell you a multidimensional story. I really enjoy the how multidimensional they are and I feel as if you have way more control in adjusting the narrative. This can be a good thing, but it can also be a bad thing. Controlling the narative can introduce bias but I didn't think that this was the case with this time lapse. It was interesting to see where the first couple cases came up and so forth. It does get a little weird as there are several datapoints, but the story is told none the less. One thing that I did find challenging was that I couldn't help but continually try to remember where I was at certain parts of the year. The date was like half a distraction to me. I do think that that may simply be specific to me though. + +## Takeaways + +I thought that with time lapses the future is for sure bright here for data science. I can't speak for all people, but I know for myself, atleast, I am almost always able to get a bigger picture if there is a narrative that comes along with it. I thought it was great that I could follow the bigger picture story progressively. Theres even more interesting things that you can do with seasons, based on the date that wasn't added here, but I do think that would be a nice additional feature. diff --git a/week4.md b/week4.md index e69de29..92528cb 100644 --- a/week4.md +++ b/week4.md @@ -0,0 +1,16 @@ +# Visualizing Country Borders Throughout History (2000 BC - 1994 AD) + +![image](https://github.com/mike3osei/reflections-1/blob/master/img/week4.gif) + +I was really intrigued by the application of this data visualization across history and time. Sometimes concepts like these are very hard to represent or explain to someone cohesively. + +## Link +https://historicborders.app/?view=globe + +## Initial Response + +My initial thoughts in regards to this visualization, were honestly how do people get the data to create such general visualizations. I thought that it was feasible, but definitely came off as if normalizing all the information would take a bit of time. To me when looking at this vis, data seems to be the challenging portion. Another thing I do wonder is what they used for the timeline. In vis where do you differentiate between a web app and a vis. I probably haven't gotten to the point where I can differentiate when it comes to web based 3d vis'. + +## Takeaways + +A lot of my takeaways this week, come in more of the form of questions. I wonder whether or not this is simply a web app? If not, what tools are used to animate these types of vis'? It was interesting to see such a dynamic vis but whats behind this stack? \ No newline at end of file diff --git a/week5.md b/week5.md index e69de29..adf7517 100644 --- a/week5.md +++ b/week5.md @@ -0,0 +1,16 @@ +# Quartetto Sincronie performing Beethoven op. 74 num. 10 + +![image](https://github.com/mike3osei/reflections-1/blob/master/img/week5.png) + +This visualization is by the Universita degli studi Roma "La Sapienza". The goal of the vis was to narrate a stage performance by Quartetto Sincronie. + +## Link +https://www.informationisbeautifulawards.com/showcase/2809-quartetto-sincronie-performing-beethoven-op-74-num-10 + +## Initial Response + +My initial thoughts in regards to this visualization were that it was pretty powerful that we could take something as confusing as a symphony performed by an orchestra and compress it to one minimal form factor. When I initially took this class, I had it in the back of my mind that I wanted to see the relationship between the arts (music in this example) and data vis. This vis, did a good job of representing amplitude in this way. I felt that I could understand the roles of different instrumentations immediately just by looking at the vis. This vis does a great job of using color as well. As someone who's color blind the colors were not too distracting or too overbearing. Even though I couldn't understand the language by simply looking at the iconography and amplitudes of the sound waves. I could immediately tell that a story was being told. + +## Takeaways + +One of my takeaways from this visualization was that it made great use of the gestalts principles of design. For me, these principles were very apparent. An example of this is the principle of proximity. When you look at the placement of elements within the vis, I immediately could testify that there was a pleasing form of cohesiveness. \ No newline at end of file diff --git a/week6.md b/week6.md index e69de29..2643f83 100644 --- a/week6.md +++ b/week6.md @@ -0,0 +1,16 @@ +# Fondation Follereau Projects Review: 2018 & 2019 + +![image](https://github.com/mike3osei/reflections-1/blob/master/img/week6.jpeg) + +This visualization is by the fondation follereau. The goal of the vis was to promote the quality of life of vulnerable African communitites. They showcase the countries budget size vs. their health projects, education projects, protection projects, and emergency projects. On the highest layer of the tree they discuss the major projects in the region. + +## Link +https://public.tableau.com/profile/noel.joe.basil#!/vizhome/Book2_16115277202920/VFSG_FF + +## Initial Response + +My initial thoughts in regards to this visualization were that it was pretty powerful that we could interact with such a large project. I know it was mentioned by the contributors that this project took a lot of funding to actually make happen. With that being said its pretty amazing to see all the entire picture as you go up the tree of specific countries. I have not really explored this approach before yet. I thought it was also interesting that the mentioned their conclusion as well. Showcasing practical ways to help and the future of new iterrations of this vis. I thought this was very interesting. + +## Takeaways + +One of my takeaways from this visualization was that it made great use of branding as a storytelling medium. Branding in this vis was great for making sense of what was going on in the vis. I thought it was great that they were also able to add illustrations of the people from the countries in their native garments around the vis. I thought it was all well thought out. \ No newline at end of file diff --git a/week7.md b/week7.md deleted file mode 100644 index e69de29..0000000 diff --git a/week8.md b/week8.md new file mode 100644 index 0000000..ebc0e1b --- /dev/null +++ b/week8.md @@ -0,0 +1,16 @@ +# ggmap: Spatial Visualization with ggplot2 + +![image](https://github.com/mike3osei/reflections-1/blob/master/img/week8.png) + +This publication describes the evolution of spacial data analysis in the Data Vis space. David Kahle and Hadley Wickham go through the creation of the ggmap tool. They describe how it can be used with open source APIs such as Google Geocoding, Distance Matrix, and Directions APIs. + +## Link +https://pdfs.semanticscholar.org/79da/0d9d7d828169db3084024a4acf6c259d0c74.pdf + +## Initial Response + +Intitially I found it very intersting as a Computer Science student how different data vis tools are created. Maybe not how they are created but what can go behind inspiring the new adoption or creation of a datavis tool. Like most innovation, creative ideas come from challenging task out there in the world and trying to simplify such things. This paper describes how one could evolve ggplot2 with the introduction of ggmap. They describe the challenges of spacial data being visualized in R, which is interesting given how one technology can be good for one thing but maybe not good for another. + +## Takeaways + +I thought it was interesting how with the introduction of a tool like ggmap, you can see other tools be generated based on the inital pain the was felt in the original R environment. Like they used that crux as a means of doing more. I feel like I can for sure take this away in my work in the future. Another takeaway with tools, is that less is always more. Creating a function with limited inputs but enough to create graphs quickly was very important in this example. \ No newline at end of file diff --git a/week9.md b/week9.md new file mode 100644 index 0000000..feaef60 --- /dev/null +++ b/week9.md @@ -0,0 +1,17 @@ +# Using Visualizations to Monitor Changes and Harvest Insights from a Global-Scale Logging Infrastructure at Twitter + +![image](https://github.com/mike3osei/reflections-1/blob/master/img/week9.png) + +This publication describes one of Twitter's largest datasets in their organization. This visualization is a unified logging infastructure that showcases user events across all their clients (mobile, web, etc.). They focus on monitoring and exploring a large collection of log events as well as how they go about performing analysis at scale. + +## Link +https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=7042487 + + +## Initial Response + +Intitially I found it very intersting how Twitter had been collecting data in their "Hadoop" data warehouse, but due to its slowness they wanted to build a whole new infastructure to perform analysis much faster. It was interesting to see that previously they utilized some open source software like MapReduce to make things easier previously but resulted on restructuring their data culture in order to take away better insights. I found it neat that no matter how big of the dataset they had everything was pretty much a tree in some way at the end of the day. I enjoyed how they went through A/B testing to figure out how users could also intertact with this data vis more efficiently instead of having to know for instance the six-part event names that define the hierarchy or domain they are interested in. + +## Takeaways + +I thought it was interesting how they chose to represent event volumes using color in the data vis. I honestly think this was the only like bad part in terms of like accessibility. You would think there would be another like variant to solidify that color and something something else was used to showcase that metric. For example using color and shape but it was pretty much just color variances that showcased the event volumes. I also did take away their testing approach when it came to improving the UX involved in datavis. I found that interesting how they were able to leverage that space to improve the quality of this data visualation. \ No newline at end of file