Missing var in census tracts 2010 v003 at v006

[jhoc1997]

The Census Tracts Basico table has a null variable in at least 40% of the income observations.

[rafapereirabr]

Hi @jhoc1997 . Could you please provide a the code you used so I replicate the issue ?

[jhoc1997]

df <- read_tracts(
year = 2010,
dataset = "Basico",
as_data_frame = TRUE
)

[rafapereirabr]

reprex for your code

library(censobr)
library(dplyr)

df <- read_tracts(
  year = 2010,
  dataset = "Basico",
  as_data_frame = TRUE
  ) |> 
  collect()


df$V003 |> summary()
#    Min. 1st Qu.  Median    Mean 3rd Qu.    Max.    NA's 
#  1.000   2.960   3.180   3.187   3.410  54.000  181174 

df$V006 |> summary()
#       Min.   1st Qu.    Median      Mean   3rd Qu.      Max.      NA's 
# 0.000e+00 3.749e+05 9.285e+05 8.830e+06 2.846e+06 1.125e+10    180370 

[rafapereirabr]

We'll look into this soon. But please make sure you're looking at the right variables. The variables v003 and v006 from the "Basico" data set are not related to income. See p.46 here:

censobr::data_dictionary(year = 2010, dataset = "tracts")

[jhoc1997]

ok, you are right. I mean V005. So, the most observation are NA's. If I try collapse by codi_muni, I will have only 3337 observation than 5565 observation, because the most of observation are NA.

[rafapereirabr]

Hi @jhoc1997 , V005 brings info on the average income of the household heads of census tracts. It is expected that some census tracts will have this as NA, for example census tracts with "collective households" (e.g. prisons) and census tracts without population.

However, you are correct there is something odd with this variable. There are too many entire municipalities with NA. Code for reprex below. We'll have a closer look into this.

library(censobr)
library(dplyr)

df <- read_tracts(
  year = 2010,
  dataset = "Basico",
  as_data_frame = TRUE
)

test <- df |> 
  group_by(code_muni) |>
  summarise(
    avg_income_head = weighted.mean(V005,w = V001, na.rm = T)
    )

head(test)

summary(test$avg_income_head)
 # Min. 1st Qu.  Median    Mean 3rd Qu.    Max.    NA's 
 #  0.0   538.0   836.5   878.9  1087.4 12043.5    1788 

Cc @antrologos , any insights here ?