final CTW project information

.gitignore

@@ -0,0 +1,3 @@
+nispuf14.dat
+.ipynb_checkpoints/
+*.DS_Store

FinalReflection.md

@@ -0,0 +1,17 @@
+# CTW Project Final Reflection
+
+
+## Backstory
+We as a team had many lofty ideas when we initially started this project, however things definitely did not work out as well as expected. A large pain point that we discovered about halfway through this project was that the initial dataset we were working on was difficult to properly clean and setup, as it required a SAS (proprietary data-science software for businesses) component that we were unable to properly utilize. As such, right before the mid-project checkin, we pivoted. We switched to a new dataset (the WHO data on mortality rates and such) and attempted to pickup right where we left off. As such, we probably did not accomplish all of the initial goals we set out for, but we felt that we were able to accomplish enough to educate others and explain a fairly large dataset to any individual.
+
+## Assessment Evidence and Interpretation
+Our main deliverable can be viewed in our repo (Jupyter notebook called Final Overview of CTW Project). This document contains the cleaning functions that we created, along with mapping functions and other graphing tools. We have 3 distinct types of graphs: bar graphs showing the number of people alive in an age group vs. the number of people dying in that same age group, life expectancies across time by age for individual regions, and the probability of dying specific to certain age brackets. We also attempted to develop a world map showing different statistics in relation to their country, however this functionality had many hiccups (the shapefile we originally needed for world data was unavailable, along with the new shapefile being corrupt after a certain number of entries). In the end, we were unable to complete a proper map, however our other graphs were fully functional and available in the notebook. These graphs show primarily age and region, in relation to the numeric values for living/dying/expectancies. This notebook provides the most full-picture idea of what we worked on, however a fourth and final graph type we used was a scatter plot consisting of life expectancies across time by age, in a .gif format for a different way to visualize aspects of the data. This dynamic visualization  has more impact in grabbing an audience’s attention, along with helping fulfill our goal of different types of visualization and exploration into that space. 
+
+## Changing the World
+In total, we feel our project gets partway to changing the world. If we were to continue, we would need to break the statistics down into countries or some breakdown more specific than “regions.” We realize this may mean we would need a more complete, or broader, dataset. A lot of the computing has already been done for the dataset we used, which means much of our work was done with post-processing data. We would need a broken down dataset because the Americas contain 1st and 3rd world countries that reasonably shouldn’t be grouped together, among other similar issues. If the data is better separated,  we could see making similar visuals that show how much longer Europeans, North Americans, and Australians can expect to live compared to someone in Africa or SouthEast Asia. Changing the world is about finding the actual shocking differences (for example, how many years lost in life expectancy) in hopes of educating and motivating individuals to join initiatives.
+
+## Learning Goals
+### Mack:
+I did learn things I wanted to know in this project. I gained comfort with matplotlib as a visualization tool and feel comfortable creating different kinds of visualizations, 3D and 2D. I think I was successful because I spent a majority of my time creating failed graphs that taught me how these tools were supposed to be used, how I was using them incorrectly, and what tools were correct for representing data the way I wanted. Unfortunately I deleted most of these early representations before I realized they were a sign of progress. At the end of this project, I was able to confidently clean up our code to be replicable, readable, and functional.
+### Philip:
+I feel that I did learn quite a bit, but it was less in the strain of visualization and more in the experiment/fail/experiment again space. Similar to Mack, many of the things I worked on did not fully pan out (looking specifically at the mapping functions, along with the foray into SAS and that failure), however I felt that any step, even in the wrong direction, was teaching me a bit more about the intricacies of matplotlib or python in the data science context. Along with that, I really enjoyed learning a lot about playing with a fairly unformatted dataset. The WHO had their own dataset standards, and when we got to making it readable for our implementation, we had to play around quite a bit to make everything work. Cleaning and formatting data was a new challenge, but it was up to us what we wanted to focus on, which made it more interesting to me in the end.