Author: Swaminathan Ramesh Sashi
1. Introduction
When you move into a new city, do you wonder which area is the best for living? In this article, I use data visualization techniques to help a family find the best area to live in Seattle. Although other cities can also be analyzed in a similar way, Seattle is chosen as an example because of ready availability of public datasets.
2. Analysis
2.1 Software used
Tableau 9.2, Excel, and R.
2.2 Dataset
The dataset is built from multiple sources. These sources are excel sheets containing various bits of data that make up the problem set.
2.3 Approaches
In order to choose the right area to live in, recreation is an important consideration. Recreational facilities in Seattle are plotted on the map with its corresponding geo-location in Graph 1. The results depicts an even distribution of recreation areas on the map except the bay area, which has a much higher density of recreation facilities.
Graph 1: Recreation Zones
Another factor to consider is the traffic congestion present in the city. Graph 2 shows traffic congestion at different areas in the city and the magnitude is denoted by the size of the circles. It seems a prudent choice to avoid high congestion areas shown on the map. High congestion counts negatively in the consideration for possible areas of residence.
Graph 2: Traffic Congestion
Crime plays an important role in selecting areas to reside. Data for certain crimes are chosen and plotted on the map as shown in Graph 3. Different colors are used for different districts, and the size of circles represents the number of crime recordings. High crime rate counts negatively towards consideration of possible areas of residence.
Graph 3: Crime Density
New housing permits are hard to obtain in big cities and budget housing is a huge problem. I tried to see if permits were issued for construction and found that most of them were for houses between 250,000 to 500,000$ as shown in Graph4.
Educational institutions are found all over the city of Seattle as shown in the Graph 5. Thus this does not play a major role in choosing areas for living.
Graph 5: Educational Institutions
Places are also checked for connectivity and the major route for travel is shown in Graph 6. Living somewhere close to these lines of transportation is the best way to commute through the city.
Graph 6: Transport Lines
2.3 Results and Summary
Considering all the factors mentioned above, I found a place to live in as shown in the map below. This area had fairly good transportation facilities, housing and education facilities and had little crime.
3. Dataset
https://data.seattle.gov/
This is very interesting. Great job. I got a question: is there a correlation between housing prices and the area that is considered more “livable’? I am curious because the price of house or rent is also a big factor for how people choose the area to live in.
I think this is a very useful application of statistical techniques to find an ideal place to live. However, I am curious how you went about weighting the different categories and combining the data to find an optimal location or whether you simply noticed patterns between the graphs. Additionally, I think adding price like SeungMin mentioned or adding housing availability could also be beneficial for making the decision.
This is very useful when considering to live in an area, especially if it is new to you. In a class I am taking we learned about principal component analysis (PCA) that analyzes large amounts of data and variables and creates principal components that are the main differentiators of the data. I wonder if this could be applied to this case to find what factors are correlated with each other and help narrow down your choices quicker.
This is a very well done analysis. Going off of what Seung Min mentioned, I would be very interested to see how each of these variables are correlated. For example, I would expect crime density and the cost of housing permits to have some degree of negative correlation. Additionally, I think including a variable for population density could be helpful in making the decision of where to live.
Im sure employment opportunities will also play a big role
I do agree with Bukola, job opportunity is also very important, and people could not live too far from where they work. And I have one more suggestion. When you deal with data about crimes, you drag both “district/sector” and “offense type” to “Color” on the Marks Card. But I think the different colors only represent different districts but not offense type, and the boundary and result are a bit confusing. Maybe you could set district color as ground color, then use different color to represent different types of crime, and size to represent sum of one certain crime.
Great post! Very interesting use of data visualization in a personal, rather than business, application. From the perspective of software development, I wonder if your analysis could be implemented into a website such as apartments.com or something similar to aid people considering relocating to a new area. Or, maybe this could be developed into an application that a company can offer their employees who are considering relocation within the company. Given that you had sufficient access to public datasets, the user could input a city of interest and a dashboard summarizing your analysis could help with their decision.
This is a very cool way to see where to live in a city. Is there a way you could layer all of these factors and visualize the best place to live in the city? im thinking an area close to public transportation, schools, and a decent price would be optimal. It would be interesting to have a heat mapof the best areas.
This was interesting for me, especially as I’m starting to look for an apartment in a new city for next year. One comment that I had is that traffic congestion in an area that you live may not be a bad thing and could actually be a good thing. I say this because the dense downtown areas near the central business district tend to have the most congestion, but they also tend to be where the jobs are. Therefore, if you lived in this congested area you might be able to avoid the congestion because you are close enough to walk or bike to work or you have a short drive. If you live in an area of low congestion it’s likely that you still have to drive to/through the areas of congestion.
Another comment is that I think in practice, deciding where to live is an interesting example of combining quantitative and qualitative factors. Some neighborhoods are just more fun, interesting, and beautiful than others, which is hard to quantify. But I bet there are some metrics that can track these factors such as density of small businesses and restaurants, the proportion of residents that are 18-30 years old, and the density of recreational facilities (which you included).
This post was very useful and it takes into account the proper metrics to perform the analysis. I agree with other comments that it would be useful to include the apartments pricing and also their disponibility. For example, it is very difficult to find a new construction in the center of an old city, thus it is also very expensive. It would be a negative aspect to consider for centric areas where the analysis shows that life is better.
Another metric could be polution or noise contamination which is probably higher in crowded or industrial areas. Finally, it is clear than the interpretation of this results is very dependent of personalpreferences since thera are people who prefer to live in a quiet place whereas others prefer to have a bussy life style.
Very interesting analysis! I like how you included the visualization of crime in each neighborhood. I didn’t even consider that a factor but now realize how important that is when choosing where to live. I would’ve liked to see the section on the housing permits explained a little more because I don’t know what a housing permit is or why it is important to live in an area where it is easy to obtain one.
This is a very interesting project. Having lived in Seattle for a summer, it seems the data confirms many of the preconceived notions. I will say, however, that the traffic visual seems to suggest that traffic is not too bad in the city when in reality, it is horrible. I am wondering what metric was used and if there is a better choice.
This is a great analysis! I did not think about using data to determine places to live. I would love to be able to implement this in determining where I would like to relocate after college. Another feature that would be interesting to compare would be cost of living in different parts of the same area/city.
I find this really interesting because I’ve had to live in two Metropolitan areas during internships that weren’t where I grew up. I think what would have been perhaps more interesting is to take a tiered view of the analysis in terms of age group. That would involve thinking about what different age groups look at in where they want to live. This could include a young professionals age group that would be really cool to read. For example, I don’t care a ton about schools when I’m thinking about where I want to live.
This is a very interesting way to visualize data. I think an additional aspect that could be added to the educational institution breakdown is the quality of schools. In my short time working in industry, it seems there are many reasons that families choose to live far away from their employment, but in my experience one of the most common reasons is because of the quality of schools. It seems this factor plays a big role for individuals who have children that are in critical stages of their educational development.
Thank you for sharing your analysis. I thought this was a useful application of data analysis techniques for a big life decision. As mentioned in some previous comments, I am a bit confused on the weight that was given to each category and how you came to the ideal conclusion. Furthermore, I would be more curious to see the types of crimes that were being committed in addition to the amount of times they were committed since some crimes are more severe than others and this would have a large impact on my decision. Depending on if you already had a job, distance to work could also be a large player as a long commute to work can be tiresome. Overall, this was interesting to see what factors you considered most important and helpful as I start to think where I would like to live.
Thank you for sharing your analysis. The visualization tool is really good. I was wondering if multiple filters can be applied all at once. Like filtering the educational institutes as well as checking for connectivity.
I think this analysis was spot on in determining the most important metrics for where to live. They also did a good job of picking a dataset with readily available information like Seattle. These statistics offer good information to people who may be moving to the city and would like a visualization of better/worse areas depending on different metrics.
However, I’m a little confused by the result; where is “the map below”? The results section is also very short and doesn’t do a great job of explaining how it decided which place (wherever that place is?) is best to live in. The analysis also doesn’t go very far past visualizing fields that already existed. While the visualizations are helpful, I got less of an analysis on which area of Seattle is best to live and why and more a different way of displaying existing data without a result or conclusion.
I find these attributes for finding a great living situation to be very helpful. It would be nice to gain a little more detail from your figures thought that may show, for example, rankings for the Educational Institutions in Graph 5. I’m not sure this is possible on Tableau. Another thought of mine after reading this is how convenient a website could be for someone looking for a living condition that suits them best like how tableau displays it. Current websites usually narrow it down to specific housing that meets someone’s criteria, but the way your graphs are displayed allows someone to pinpoint on the map which specific area they would consider living, and then that person would have a much more narrowed down option of where (area) in the city they’d like to live.
As housing prices keep rising in cities like Seattle, finding a preferable place to live can be a nightmare. House shopping presents a great way to get to know a new city, but can become a daunting task easily once all factors are accounted for. Seeing visual maps for crime rates, schooling locations, recreational centers, public transportation, and street congestion has helped me get a better grasp of how the city of Seattle is operated. Having never been there myself, I would love to see a map displaying the median age of residents in each district. I think it is one of many displays that could increase the clarity of this report and ultimately help the customer make a more informed purchase.
This visualization method is really cool, and I’d like to add one thing. When trying to choose a new place to live there are various calculators online which will allow you to enter a starting location, a salary and an ending location. Then it will give you an estimated salary needed to maintain the same standard of living.
For example: https://smartasset.com/mortgage/cost-of-living-calculator#kPos8QKyQK
I found it helpful when considering job offers.
When looking for a home or apartment in a city that you have never been too, this approach is very helpful and gives people a place to start with very little investment and time. Time could be the most important if you get a new job and have to have a place to live fairly quickly.
How about social aspects such as rental and general expenses because those will be a big consideration for college grads like me.
This was a really cool analysis, Swaminathan! I found the traffic congestion analysis particularly interesting, as I think this is going to be a major factor in most cities within my lifetime, especially considering ever increasing urban populations. Although I’m sure there are companies offering this sort of analytical service, I think it’s really great that you were able to statistically determine which areas of a city have the most congestion. If you could get the corresponding times of peak congestion to create a temporal traffic density map, that would allow you to, at a glance, see the areas of each city to avoid. Maybe, if nobody offers this service already, you should think about starting a new company!
I really enjoyed the analyses you presented in this report overall, and like how you took a different approach to data analysis than many of the other project examples found on this site. Great job!
This is such an interesting and informative project! I think you have considered some really relevant factors like schools and traffic to identify which place is best to live in. I think that age can also play a big part in the type of factors that individuals consider when moving to a new city. For instance, if a recent graduate is moving to the city, they are probably not going to be much concerned about nearby schools but would want to live in an area with a lot of job opportunities. For this, it would be beneficial to combine all these factors into one heat map with filters for the factors used so that the user can select all the factors that are important to them and ultimately identify the best place to live in.
I appreciate this topic as I am starting to consider where I should live next Summer in Chicago. One other important metric that would be good to look at would be the age distribution of different regions across Seattle because people typically like to live around others who are of similar age to them. In the introduction you mentioned that you are approaching this topic as a problem for families so I understand why you would have suggested avoiding high traffic areas. However, age is important because many young individuals like myself would prefer to live in high traffic areas where there is more going on.
I thought the attributes being considered when choosing a place to live are important and not many people think about all when deciding. I think the graph regarding crime in areas was a little hard to understand and could maybe be enhanced with adding another layer.
This is super interesting, especially considering that I am a senior and will be graduating in the spring. This analysis has given me other considerations that I would not have thought of when I’m deciding on the apartment I want to live in next year, especially since it’s going to be a new city.