Spaces:
Build error
Build error
spatel54 commited on
Commit ·
9fbde92
1
Parent(s): 5d065f8
Update
Browse files
app.py
CHANGED
|
@@ -59,9 +59,6 @@ st.write("For data transformations, I removed all NaN values. The columns with N
|
|
| 59 |
st.write("### Encoding")
|
| 60 |
st.write("I encoded 'Year Acquired' as a numerical value on the X axis value and 'Agency Name' as a nominal variable Y axis value to see the buildings that were acquired over time by each agency. I did not set 'Year Acquired' as an quantative variable because the X axis would aggregate the years which would provide useless meaning to the user so I chose ordinal to list out each year. I incoporated jitter to decrease overlapping of data points through the yOffset variable. The tooltip can also show the building status, year constructed, usage description, square footage, and address. I chose building status so the user can quickly view the status of the building if the color scheme is not clear. Year constructed is included to determine age of building as this can be a factor for building inspection and maintenance. Usage description helps determine the intent of the building and helps keep track of the different types of buildings for an agency. Square footage helps determine size of building which can help determine resource allocation. Finally, address is included so the user can see the building in their browser instead of having to look at the dataset to find the address. For the color scheme, I chose 'set1' because according to the altair site, set1 is a color scheme specifically used to handle categorical variables to map discrete values such as agency names. I chose to color the data points by the 'Bldg Status' variable to easily discern the status of each building. The different shades of colors may be due to the fact that certain data points are overlapping despite jitter being applied. If multiple buildings share the same agency, year acquired, and status, this may make the dots stack and appear darker in color. Lighter dots might mean that there are less buildings that matches the same acquisition year, agency, and building status. Finally, for the '.transform_filte'r encoding parameter, I determined the value of the jitter using a mathematical formula.")
|
| 61 |
|
| 62 |
-
st.write("### Interesting Observation!")
|
| 63 |
-
st.write("I learned that the ECEB building on campus was constructed between 2011-2014. This can be easily indicated by my visualization because this building was listed as in progress (blue dot) in 2013. I was amazed by this connection! It was also interesting to see that a lot of buildings that were acquired was constructed in the same year for all of the buildings. This may due to the size of the building so construction may have been completed in under a year.")
|
| 64 |
-
|
| 65 |
st.write("### What I would change in the future")
|
| 66 |
st.write("I noticed that the colors in this visualization are red, blue and green. This color pallete is not accessibility friendly since individuals with colorblindness would have difficulty interpretating the visualization. I would use a darker green, red, and yellow for Bldg Status.")
|
| 67 |
|
|
|
|
| 59 |
st.write("### Encoding")
|
| 60 |
st.write("I encoded 'Year Acquired' as a numerical value on the X axis value and 'Agency Name' as a nominal variable Y axis value to see the buildings that were acquired over time by each agency. I did not set 'Year Acquired' as an quantative variable because the X axis would aggregate the years which would provide useless meaning to the user so I chose ordinal to list out each year. I incoporated jitter to decrease overlapping of data points through the yOffset variable. The tooltip can also show the building status, year constructed, usage description, square footage, and address. I chose building status so the user can quickly view the status of the building if the color scheme is not clear. Year constructed is included to determine age of building as this can be a factor for building inspection and maintenance. Usage description helps determine the intent of the building and helps keep track of the different types of buildings for an agency. Square footage helps determine size of building which can help determine resource allocation. Finally, address is included so the user can see the building in their browser instead of having to look at the dataset to find the address. For the color scheme, I chose 'set1' because according to the altair site, set1 is a color scheme specifically used to handle categorical variables to map discrete values such as agency names. I chose to color the data points by the 'Bldg Status' variable to easily discern the status of each building. The different shades of colors may be due to the fact that certain data points are overlapping despite jitter being applied. If multiple buildings share the same agency, year acquired, and status, this may make the dots stack and appear darker in color. Lighter dots might mean that there are less buildings that matches the same acquisition year, agency, and building status. Finally, for the '.transform_filte'r encoding parameter, I determined the value of the jitter using a mathematical formula.")
|
| 61 |
|
|
|
|
|
|
|
|
|
|
| 62 |
st.write("### What I would change in the future")
|
| 63 |
st.write("I noticed that the colors in this visualization are red, blue and green. This color pallete is not accessibility friendly since individuals with colorblindness would have difficulty interpretating the visualization. I would use a darker green, red, and yellow for Bldg Status.")
|
| 64 |
|