Week Seven of Data Science for the Public Good

Demographic Analysis Time

Week Seven
R
Demographic Analysis
Author

Kailyn Hogan

Published

June 30, 2023

Monday

On Monday, my group had our final? meeting with our stakeholders, Erin Mullinex and Erin Olson-Douglas. I got some clarification on what they expected from the demographic analysis from the meeting. They wanted me to look further into economic and fiscal conditions of Iowa’s communities.

Erin and Erin’s intention with our project is to improve the housing stock across the state. To do so, we need to identify which communities need improvement based on external factors. The final destination after three years is to create an application that assess housing conditions and generates a map of housing characteristic quality.

Erin and Erin also let me know that thriving communities tend to have an employer of some sort. Communities with an employer do well because the employer may provide grants and funding to improve the community. Thriving communities are also growing in population and sometimes have lending institutions and non-profits.

So far, I know to look at the following characteristics in the communities that could benefit from our project:

  • populations between 500 and 10,000

  • population change from 2000-2020: growing, stable, or shrinking?

  • median age of residents

  • presence of ag-related industries

  • home ownership rates

  • percent of residents commuting to work

  • number of jobs

  • median income

  • income change

  • workforce characteristics (this one is still confusing)

  • local spending patterns (don’t know where to find this data yet)

After the meeting with Erin and Erin, I spent the rest of the day sorting WINVEST photos to train our AI Models. The photos are stored in CyBox under the communities they are associated with. Chris made a CSV file of the WINVEST data that we used to filter for poor roofs, poor siding, poor landscaping, and poor gutter. The CSV has the image name in a column, so we could then manually search for the image names associated with the poor values. I also went to Durham’s GIS lab to create community summary infographics for Independence, New Hampton, and Grundy Center.

Community summary infographics are really easy to make.

  1. Download shapefile for the geography you want to make an infographic for. For me, this was the places geography that I downloaded from the Census’s shapefile directory.
  2. Insert the geography shapefile into ArcGIS Pro.
  3. Search for infographics in the tool bar.
  4. Click on the geography you want to make an infographic for.
  5. Select which type of infographic you would like. There are a lot of pre-loaded infographics available for free on ArcGIS Pro. I chose the Community Summary template to create the infographics above.

Some of the elements on the infographics you can toggle to compare to different geographies. The Iowa community summary had the option to compare to the United States. The city infographics I could choose to compare to the county, state, or nation.

Tuesday

On Tuesday, I was able to start the demographic analysis. I worked with population data and housing data from the 2021 5-Year American Community Survey.

library(tidycensus)
library(tidyverse)


## Population Data ##
# Getting population data from 2000-2020 Decennial Census.

# The 2020 total population API is "P1_001N."
# The 2010 and 2000 total population API is "P001001."
pop <- get_decennial(geography = "place",
                     state = "IA",
                     variables = c("pop20" = "P1_001N"),
                     year = 2020,
                     output = "wide")
pop10 <- get_decennial(geography = "place",
                state = "IA",
                variable = c("pop10"="P001001"),
                year = 2010,
                output = "wide")
pop00 <- get_decennial(geography = "place",
                       state = "IA",
                       variable = c("pop00"="P001001"),
                       year = 2000,
                       output = "wide")
# Join the 2020, 2010, and 2000 total population data by GEOID and NAME.
pop <- pop %>% 
  left_join(pop10,by = c("GEOID","NAME"))
pop <- pop %>% 
  left_join(pop00,by = c("GEOID","NAME"))

# Next, determine which cities are growing, shrinking, or stable in population.
# Will need to calculate the percent population change (first pop - second pop / first pop).
# A stable population is between -2% and 2% population change.
# An increasing population has a greater than 2% population change.
# A decreasing population has a less than -2% population change. 
pop <- pop %>%
  mutate(prc_change = ((pop20 - pop00) / pop20)) %>% #calculating percent change over 2000-2020
  mutate(change_label = ifelse(prc_change > .02, "Growing",
                             ifelse(prc_change < -.02, "Shrinking", "Stable")))

head(pop)
GEOID NAME pop20 pop10 pop00 prc_change change_lable
1900190 Ackley 1599 1589 1809 -0.131332083 Shrinking
1900235 Ackworth 115 83 85 0.260869565 Growing
1900370 Adair 791 781 839 -0.060682680 Shrinking
1900505 Adel 6153 3682 3435 0.441735739 Growing
1900595 Afton 874 845 917 -0.049199085 Shrinking 
## Housing Data ##
# Getting housing data from the 2021 5-Year American Community Survey.

# Using the following API codes from the ACS:
# total housing units = "B25001_001"
# owner occupied units = "B25003_002"
# renter occupied units = "B25003_003"
# total occupied units = "B25002_002"
# total vacant units = "B25002_003"
# median house value = "B25077_001" 
# median house age = "B25035_001"

housing <- get_acs(geography = "place",
                   state = "IA",
                   variable = c("total_units" = "B25001_001",
                                "occupied_units" = "B25002_002",
                                "vacant_units" = "B25002_003",
                                "owner_occupied" = "B25003_002",
                                "renter_occupied" = "B25003_003",
                                "median_house_value" = "B25007_001"),
                   year = 2021,
                   output = "wide")
# Getting median year built data from ACS.
median_age <- get_acs(geography = "place",
                      state = "IA",
                      variable = c("median_year_built" = "B25035_001"),
                      year = 2021,
                      output = "wide") 
# Calculate the median house age by subtracting the median year built from 2023.
median_age <- median_age %>%
  mutate(median_year_builtE = ifelse(median_year_builtE == 0, NA, median_year_builtE)) %>%
  mutate(median_house_ageE = 2023 - median_year_builtE) %>% 
  mutate(median_house_ageM = median_year_builtM) # Don't have to make a new moe because the subtraction didn't introduce new errors to the data.

# Join the median year built data to the larger housing data frame.
housing <- housing %>% 
  left_join(median_age, by = c("GEOID","NAME"))

# Calculate home ownership, vacany, and rental rates.
# ALL RATES ARE PERCENTAGES with these calculations.
housing <- housing %>% 
  mutate(home_ownership_rateE = (owner_occupiedE / occupied_unitsE)) %>%  # Divide owner occupied units by total occupied units.
  mutate(home_ownership_rateM = (sqrt((owner_occupiedM^2) / (occupied_unitsE^2) + ((owner_occupiedE * occupied_unitsM)^2) / (occupied_unitsE^4)))) %>%  # Calculate the new moe.
  mutate(rental_rateE = (renter_occupiedE / occupied_unitsE)) %>%  # Divide renter occupied units by total occupied units.
  mutate(rental_rateM = (sqrt((renter_occupiedM^2) / (occupied_unitsE^2) + ((renter_occupiedE * occupied_unitsM)^2) / (occupied_unitsE^4)))) %>% # Calculate the new moe.
  mutate(vacancy_rateE = (vacant_unitsE / total_unitsE)) %>%  # Divide vacant units by total units.
  mutate(vacancy_rateM = (sqrt((vacant_unitsM^2) / (total_unitsE^2) + ((vacant_unitsE * total_unitsM)^2) / (total_unitsE^4)))) # Calculate the new moe.

head(housing)
GEOID NAME total_unitsE total_unitsM occupied_unitsE occupied_unitsM vacant_unitsE vacant_uintsM owner_occupiedE owner_occupiedM renter_occupiedE renter_occupiedM median_house_valueE median_house_valueM median_year_builtE median_year_builtM median_house_ageE median_house_ageM home_ownership_rateE home_ownership_rateM vacancy_rateE vacancy_rateM rental_rateE rental_rateM
1900190 Ackley 800 71 705 69 95 39 553 59 152 48 705 69 1955 5 68 5 0.7843972 0.11356683 0.118750000 0.049876190 0.21560284 0.07128013
1900235 Ackworth 47 20 47 20 0 10 35 17 12 11 47 20 1999 11 24 11 0.7446809 0.48087923 0.000000000 0.212765957 0.25531915 0.25803094
1900370 Adair 384 60 344 59 40 27 207 44 137 44 344 59 1960 8 63 8 0.6017442 0.16435236 0.104166667 0.072171720 0.39825581 0.14500288
1900505 Adel 2312 151 2234 120 78 97 1548 171 686 157 2234 120 1978 7 45 7 0.6929275 0.08511417 0.033737024 0.042012837 0.30707252 0.07218725
1900595 Afton 382 56 357 53 25 16 251 40 106 36 357 53 1962 3 61 3 0.7030812 0.15313072 0.065445026 0.042969563 0.29691877 0.11005386 
## Taxable Property Values ##
#link to data: https://data.iowa.gov/Local-Government-Finance/Taxable-Property-Values-in-Iowa-by-Tax-District-an/ig9g-pba5

taxable.csv <- read.csv("C:/Users/Kailyn Hogan/OneDrive - Iowa State University/Documents/GitHub/Housing/demographics analysis/Community Profile Datasets/Taxable_Property_Values_in_Iowa_by_Tax_District_and_Year.csv")
# The name of the city is stored in City.Name and is in uppercase.
# For the Census data, only the first letter of the city name is capitalized and " city, Iowa|, Iowa" is attached.
# Need to lowercase taxable property values City.Name and remove " city, Iowa|, Iowa" from housing and population data frames before joining them. 

housing <- housing %>% 
  mutate(NAME = str_remove(NAME," city, Iowa|, Iowa"))
pop <- pop %>% 
  mutate(NAME = str_remove(NAME, " city, Iowa|, Iowa"))
taxable.csv <- taxable.csv %>% 
  mutate(City.Name = str_to_sentence(City.Name)) #str_to_sentence() uses regular sentence formatting where the first letter is capitalized.
                              # Could have also used str_to_titletext().

# Aggregate the sum of all columns and group by City.Name.
# Function is set to sum to get the sum of all values.
# na.rm = TRUE means the NAs get ignored when summing.
residential <- aggregate(Residential ~ City.Name, data = taxable.csv, FUN = sum, na.rm = TRUE)
ag_land <- aggregate(Ag.Land ~ City.Name, taxable.csv, FUN = sum, na.rm = TRUE)
ag_building <- aggregate(Ag.Building ~ City.Name, taxable.csv, FUN = sum, na.rm = TRUE)
commercial <- aggregate(Commercial ~ City.Name, taxable.csv, FUN = sum, na.rm = TRUE)
industrial <- aggregate(Industrial ~ City.Name, taxable.csv, FUN = sum, na.rm = TRUE)
# Join aggregated dataframes together to reform taxable property values data frame.
taxable_prop_values <- residential %>%
  left_join(ag_land, by = "City.Name") %>%
  left_join(ag_building, by = "City.Name") %>%
  left_join(commercial, by = "City.Name") %>%
  left_join(industrial, by = "City.Name")
  
head(taxable_property_values)
City.Name Residential Ag.Land Ag.Building Commercial Industrial
57582069434 68289755861 3756099698 6892181005 9990249878
Ackley 59077903 1756505 61418 20369025 2143449
Ackworth 10399627 403117 7779 67230 0
Adair 1151067 715063 40676 5091781 195604
Adel 257984582 2651441 27555 81299747 12757527
Afton 24900696 138605 38906 7514181 1583244

I also download the LEHD Origin-Destination Employment Statistics (LODES) from the Census website as well. I downloaded the 2020 Residential Area Characteristics and the 2020 Workplace Area Characteristics. They are aggregated by census block. I downloaded the Geography crosswalk for IA as well. The geography is needed to associate the census block codes to their related places in Iowa.

Wednesday

Late on Wednesday, my team got more clarification from our boss, Chris, on what the AI Housing project actually is. From our conversation, I interpreted that our project is intended to produces a rapid, hands off, housing quality assessment for policy and grant decision makers.

## Median Age of Residents ##

# Getting the median age of all people in all places in Iowa from 2021 5-Year American Community Survey.
pop_age <- get_acs(geography = "place",
               state = "IA",
               year = 2021,
               variable = c("med_age" = "B01002_001"),
               output = "wide") %>% 
  mutate(NAME = str_remove(NAME, " city, Iowa|, Iowa"))

# Get the number of people under 18 and over 65 for all places in Iowa.
# The age data in the American Community Survey is separated by sex, so you will need to get the data for men and then women and combine the data frames. 
male <- get_acs(geography = "place",
                  state = "IA",
                  year = 2021,
                  variable = c("under5" = "B01001_003", # Make sure to start the variable names with a letter.
                               "a5to9" = "B01001_004",
                               "a10to14" = "B01001_005",
                               "a15to17" = "B01001_006",
                               "a65to66" = "B01001_020",
                               "a67to69" = "B01001_021",
                               "a70to74" = "B01001_022",
                               "a75to79" = "B01001_023",
                               "a80to84" = "B01001_024",
                               "over85" = "B01001_025",
                               "total" = "B01001_002"),
                output = "wide") %>% 
  mutate(NAME = str_remove(NAME, " city, Iowa|, Iowa"))
male <- male %>% 
  mutate(under18E = under5E + a5to9E + a10to14E + a15to17E) %>%
  mutate(under18M = sqrt(under5M^2 + a5to9M^2 + a10to14M^2 + a15to17M^2)) %>%
  #mutate(prc_under18E = under18E / totalE) %>% 
  #mutate(prc_under18M = moe_ratio(under18E, totalE, under18M, totalM)) %>% 
  mutate(over65E = a65to66E + a67to69E + a70to74E + a75to79E + a80to84E + over85E) %>%
  mutate(over65M =  sqrt(a65to66M^2 + a67to69M^2 + a70to74M^2 + a75to79M^2 + a80to84M^2 + over85M^2)) %>% 
 # mutate(prc_over65E = over65E / totalE) %>% 
  #mutate(prc_over65M = moe_ratio(over65E, totalE, over65M, totalM)) 

# Now do the same for the female age data.
female <- get_acs(geography = "place",
                state = "IA",
                year = 2021,
                variable = c("under5" = "B01001_027",
                             "a5to9" = "B01001_028",
                             "a10to14" = "B01001_029",
                             "a15to17" = "B01001_030",
                             "a65to66" = "B01001_044",
                             "a67to69" = "B01001_045",
                             "a70to74" = "B01001_046",
                             "a75to79" = "B01001_047",
                             "a80to84" = "B01001_048",
                             "over85" = "B01001_049",
                             "total" = "B01001_026"),
                output = "wide") %>% 
  mutate(NAME = str_remove(NAME, " city, Iowa|, Iowa"))
female <- female %>% 
  mutate(under18E = under5E + a5to9E + a10to14E + a15to17E) %>%
  mutate(under18M = sqrt(under5M^2 + a5to9M^2 + a10to14M^2 + a15to17M^2)) %>%
  #mutate(prc_under18E = under18E / totalE) %>% 
 #mutate(prc_under18M = moe_ratio(under18E, totalE, under18M, totalM)) %>% 
  mutate(over65E = a65to66E + a67to69E + a70to74E + a75to79E + a80to84E + over85E) %>%
  mutate(over65M =  sqrt(a65to66M^2 + a67to69M^2 + a70to74M^2 + a75to79M^2 + a80to84M^2 + over85M^2)) %>% 
  #mutate(prc_over65E = over65E / totalE) %>% 
  #mutate(prc_over65M = moe_ratio(over65E, totalE, over65M, totalM)) 

# Combine the data frames.
age <- female %>% 
  bind_rows(male)

# Aggregate the data by NAME.
# I am using a different aggregation format here by using the summary() function. 
# This will produce a separate dataframe with just under18E, under18M, over65E, over65M, prc_under18E, prc_under18M, prc_over65E, and prc_over65M.
aggregated_age <- age %>% 
  group_by(NAME) %>% 
             summarize(under18E = sum(under18E),
                       under18M = sqrt(sum(under18M^2)),
                       over65E = sum(over65E),
                       over65M = sqrt(sum(over65M)),
                       prc_under18E = sum(under18E)/sum(totalE),
                       prc_under18M = sqrt(sum(under18M^2) / sum(totalE)^2 + (sum(under18E)^2 * sum(totalM^2)) / sum(totalE)^4),
                       prc_over65E = sum(over65E)/sum(totalE),
                       prc_over65M = sqrt(sum(over65M^2) / sum(totalE)^2 + (sum(over65E)^2 * sum(totalM^2)) / sum(totalE)^4))

# Add a column that states whether a place is "aged", "stable" or "young."
# Using a 2% difference to gauge a stable population.
# Aged population has a less than -2% difference between the percent under 18 and the percent over 65.
# Young population has a greater than 2% difference between the percent under 18 and the percent over 65. 
aggregated_age <- aggregated_age %>%
  mutate(age_label = ifelse(prc_under18E - prc_over65E > .02, "Young",
                               ifelse(prc_under18E - prc_over65E < -.02, "Aging", "Stable")))

## Add the age data to the population dataframe.
pop <- pop %>% 
  left_join(aggregated_age,by = c("NAME")) %>% 
  left_join(pop_age, by = c("GEOID","NAME"))

head(aggregated_age)
NAME under18E under18M over65E over65M prc_under18E prc_under18M prc_over65E prc_over65M age_label
Ackley 402 68.39591 349 9.030714 0.23563892 0.044036378 0.20457210 0.0166781814 Young
Ackworth 50 26.47640 24 6.256889 0.33783784 0.212313083 0.16216216 0.0692779288 Young
Adair 206 67.47592 144 7.692109 0.24065421 0.088939653 0.16822430 0.0301622637 Young
Adel 1708 285.81987 661 14.821549 0.29111982 0.050896612 0.11266405 0.0062375436 Young
Afton 208 60.94260 189 8.606830 0.22757112 0.073795265 0.20678337 0.0302370669 Young 
library(tidyverse)

## Unemployment ##
# Getting unemployment data from the 2021 5-Year American Community Survey.

# Get the total number of people in the workforce and the total number of the labor force that is unemployed.
# Calculate the percent of the workforce that is unemployed. 
unemployment <- get_acs(state = "IA", 
                   geography = "place",
                   year = 2021,
                   variable = c("total_workforce" = "B23025_003",
                                "unemployed" = "B23025_005"),
                   output = "wide") %>% 
  mutate(prc_unemployed = unemployedE / total_workforceE,
         unemployed_moe = moe_ratio(unemployedE, total_workforceE, unemployedM, total_workforceM)) %>% 
  mutate(NAME = str_remove(NAME," city, Iowa|, Iowa"))

head(unemployment)
GEOID NAME total_workforceE total_workforceM unemployedE unemployedM prc_unemplyed unemployed_moe
1900190 Ackley 828 88 25 17 0.030193237 0.020780658
1900235 Ackworth 73 39 6 9 0.082191781 0.130873976
1900370 Adair 440 101 25 25 0.056818182 0.058295873
1900505 Adel 3125 245 127 93 0.040640000 0.029930074
1900595 Afton 480 100 42 31 0.087500000 0.067106702 
## Commuting ##
library(tidyverse)
library(tidycensus)

# Getting commute data from the 2021 5-Year American Community Survey.

# Get the total number of people available for commute data and the total number of people commuting to work. 
# Calculate the percent of people commuting to work. 
prc_travel <- get_acs(geography = "place",
                     state = "IA",
                     variable = c("total" = "B08008_001",
                                  "travel" = "B08008_004"),
                     year = 2021,
                     output = "wide") %>% 
  mutate(prc_travel = travelE / totalE,
         travel_moe = (moe_ratio(travelE, totalE, travelM, totalM))) %>% 
  mutate(NAME = str_remove(NAME, " city, Iowa|, Iowa"))

head(prc_travel)
GEOID NAME totalE totalM travelE travelM prc_travel travel_moe
1900190 Ackley 797 85 534 83 0.6700125 0.12629855
1900235 Ackworth 67 38 64 39 0.9552239 0.79519912
1900370 Adair 412 93 230 57 0.5582524 0.18713615
1900505 Adel 2924 229 2216 263 0.7578659 0.10776390
1900595 Afton 434 90 319 77 0.7350230 0.23390328

When I was working on the demographic analysis on Wednesday, I took a break to do some more research on demographic analyses in general. I found this link on my project leader’s, Liesl Eathington, ISU webpage.

The link is to a local news article that Liesl was interviewed for talking about the characteristics of thriving rural, small towns. The article also talked about some demographic analysis that was done on rural communities for grant purposes. Through this analysis, the researchers discovered that it was key to analyze housing, income, and jobs in a community to understand it. The researchers also found that an indicator of a declining small town is deteriorating infrastructure like streets and sewers.

Small towns that are thriving have one or more of the following characteristics:

  • close healthcare

  • local day care options

  • local grocery store

  • local utilities

  • local restaurants

  • local shops

  • local school(s)

Knowing this information will help me finish the demographic analysis of Iowa’s communities. I want to see if I can find available data on infrastructure conditions to identify communities potentially in need of our AI Model.

Thursday

My group spent all morning working on our team’s teaser video. I worked on this blog and this week’s team blog when I had breaks from that. The team blog can be found on this blog and on the 2023 DSPG blog.

I compiled the data I currently have for the demographic analysis into a larger data frame called “iowa.” I used the function write.csv() to create a CSV file from the “iowa” dataframe. This still needs updated because I am not done working with the WAC and RAC data, but it is a starting point. Later, I will be putting this CSV file into Tableau to visualize the data.

library(tidyverse)
iowa # Name of aggregated database.

iowa <- pop %>% 
  left_join(housing, by = c("GEOID","NAME")) %>% 
  left_join(taxable_prop_values, by = "NAME") %>% 
  left_join(unemployment, by = c("GEOID","NAME")) %>% 
  left_join(prc_travel, by = c("GEOID","NAME"))

# Export as CSV.
write.csv(iowa, "analysing_iowa_communities.csv", row.names = FALSE)
# Read the CSV file. 
data <- read.csv("C:/Users/Kailyn Hogan/OneDrive - Iowa State University/Documents/GitHub/Housing/demographics analysis/analysing_iowa_communities.csv") 

# Display the contents of the CSV file.
knitr::kable(head(arrange(data, NAME)))
GEOID NAME pop20 pop10 pop00 prc_change change_label under18E under18M over65E over65M prc_under18E prc_under18M prc_over65E prc_over65M age_label med_ageE med_ageM total_unitsE total_unitsM occupied_unitsE occupied_unitsM vacant_unitsE vacant_unitsM owner_occupiedE owner_occupiedM renter_occupiedE renter_occupiedM median_house_valueE median_house_valueM median_year_builtE median_year_builtM median_house_ageE median_house_ageM home_ownership_rateE home_ownership_rateM vacancy_rateE vacancy_rateM rental_rateE rental_rateM Residential Ag.Land Ag.Building Commercial Industrial total_workforceE total_workforceM unemployedE unemployedM prc_unemployed unemployed_moe totalE totalM travelE travelM prc_travel travel_moe
1900190 Ackley 1599 1589 1809 -0.1313321 Shrinking 402 68.39591 349 9.030714 0.2356389 0.0440364 0.2045721 0.0166782 Young 41.5 3.0 800 71 705 69 95 39 553 59 152 48 705 69 1955 5 68 5 0.7843972 0.1135668 0.1187500 0.0498762 0.2156028 0.0712801 59077903 1756505 61418 20369025 2143449 828 88 25 17 0.0301932 0.0207807 797 85 534 83 0.6700125 0.1262985
1900235 Ackworth 115 83 85 0.2608696 Growing 50 26.47640 24 6.256889 0.3378378 0.2123131 0.1621622 0.0692779 Young 38.1 30.8 47 20 47 20 0 10 35 17 12 11 47 20 1999 11 24 11 0.7446809 0.4808792 0.0000000 0.2127660 0.2553191 0.2580309 10399627 403117 7779 67230 0 73 39 6 9 0.0821918 0.1308740 67 38 64 39 0.9552239 0.7951991
1900370 Adair 791 781 839 -0.0606827 Shrinking 206 67.47592 144 7.692109 0.2406542 0.0889397 0.1682243 0.0301623 Young 37.7 8.6 384 60 344 59 40 27 207 44 137 44 344 59 1960 8 63 8 0.6017442 0.1643524 0.1041667 0.0721717 0.3982558 0.1450029 1151067 715063 40676 5091781 195604 440 101 25 25 0.0568182 0.0582959 412 93 230 57 0.5582524 0.1871361
1900505 Adel 6153 3682 3435 0.4417357 Growing 1708 285.81987 661 14.821549 0.2911198 0.0508966 0.1126641 0.0062375 Young 36.8 1.5 2312 151 2234 120 78 97 1548 171 686 157 2234 120 1978 7 45 7 0.6929275 0.0851142 0.0337370 0.0420128 0.3070725 0.0721873 257984582 2651441 27555 81299747 12757527 3125 245 127 93 0.0406400 0.0299301 2924 229 2216 263 0.7578659 0.1077639
1900595 Afton 874 845 917 -0.0491991 Shrinking 208 60.94260 189 8.606830 0.2275711 0.0737953 0.2067834 0.0302371 Young 39.0 9.2 382 56 357 53 25 16 251 40 106 36 357 53 1962 3 61 3 0.7030812 0.1531307 0.0654450 0.0429696 0.2969188 0.1100539 24900696 138605 38906 7514181 1583244 480 100 42 31 0.0875000 0.0671067 434 90 319 77 0.7350230 0.2339033
1900640 Agency 620 638 622 -0.0032258 Stable 71 29.79933 123 6.383032 0.1501057 0.0678777 0.2600423 0.0458014 Aging 49.6 10.8 233 41 211 38 22 22 154 29 57 25 211 38 1959 6 64 6 0.7298578 0.1901772 0.0944206 0.0958713 0.2701422 0.1280830 82024799 119286 70 2742290 2872782 240 63 7 7 0.0291667 0.0301548 232 63 219 63 0.9439655 0.3734276

Friday

On Friday, I gave the weekly wrap up presentation.