In danielmarcelino/SenadoBR: Wraps the API for the Brazilian Senate

knitr::opts_chunk$set(
  collapse = TRUE,
  message = FALSE,
  warning = FALSE,
  comment = "",
  echo = FALSE,
  cache = FALSE,
  fig.align = "center",
  fig_width = 6.75,
  fig_height = 5.5,
  fig.retina = 2,
  out.width = "90%"
)

Dados

library(tidyverse)
library(data.table)
library(DT)
library(SenadoBR)
library(knitr)
options(dplyr.summarise.inform = FALSE)
options(tibble.print_max = 100)

Parlamentares

senadores = fetchSenators()


DT::datatable(senadores, filter = 'none', rownames = FALSE, options = list(pageLength = 15, autoWidth = TRUE
))

Dados de votações nominais

votacoes <- fetchRollcallVotesByInterval()

print(votacoes)

# votacoes %>% rmarkdown::paged_table()
#

Construtor de URLs para dowlnoads

votacoes <- votacoes %>% mutate(pdf_url = glue::glue("http://rl.senado.gov.br/reports/rwservlet?legis&report=/forms/parlam/vono_r01.RDF&paramform=no&p_cod_materia_i={bill_id}&p_cod_materia_f={bill_id}&p_cod_sessao_votacao_i={rollcall_id}&p_cod_sessao_votacao_f={rollcall_id}"))


glimpse(votacoes)

ids <- votacoes %>% pull(rollcall_id)


# ids <- votacoes %>% pull(url_votacao) %>% stringr::str_extract( "(?<=materia\\/).+?(?=\\/)")

rollcalls <- map_df(ids, ~{
    fetchRollcallVotes(.x)})

print(rollcalls)

urls <- votacoes %>% pull(pdf_url)

# rollcalls <- tibble::tibble()

# for(url in urls){
# dat <-fetch_votos_por_link_votacao_senado(url) 
# rollcalls <- bind_rows(rollcalls, dat)
# }


rollcalls <- map_df(urls, ~{
    fetch_all_votos_senado(.x)})

data <- 
    rollcalls %>% 
    left_join(
      senadores %>%
        select(legislator_name, legislator_id), 
      by=c("legislator_name"))

rollcalls %>%
  dplyr::count(legislator_vote)


base <- rollcalls %>%
  dplyr::distinct(bill_id, rollcall_id, legislator_name, .keep_all = TRUE)



base %>%
  dplyr::filter(bill_id == 123903, rollcall_id == 5952) %>% View()


# Limpando a base
base <- transformVotes(data, filter = FALSE)


base %>%
  # filter(!is.na(Governo)) %>%
  dplyr::distinct(rollcall_id) %>%
  nrow()

base %>%
  filter(!is.na(Governo)) %>%
  dplyr::distinct(rollcall_id) %>%
  nrow()

Governismo geral

# base %>% group_by(legislator_party) %>% summarise(n=n()) %>% data.frame()

## Governismo geral
(governismo_geral <- base %>% 
  # filter(decision_date < as.Date("2020-07-25")) %>%
  dplyr::group_by(rollcall_id) %>%
  dplyr::summarise(governismo = mean(governismo, na.rm=T),.groups = 'drop') %>%
  dplyr::summarise(governismo = mean(governismo, na.rm=T), .groups = 'drop'))


## Média de frequência nas votações
base %>%
    dplyr::group_by(rollcall_id) %>%
    dplyr::summarise(Freq = n()) %>%
    dplyr::summarise(Freq = mean(Freq))

## Frequência nas votações 30 dias
base %>%
  dplyr::filter(lubridate::ymd(decision_date) >=    lubridate::today() - 30) %>%
  dplyr::group_by(rollcall_id) %>%
  dplyr::summarise(Freq = n()) %>%
  dplyr::summarise(Freq = mean(Freq))



# Frequência média de parlamentares com governismo acima de 80%
governismo_legislator <- base %>%
  dplyr::group_by(rollcall_id, legislator_id , legislator_name) %>%
  dplyr::summarise(soma = sum(governismo, na.rm=TRUE),
            nao = sum(abs(-1 + governismo)),
            governismo = mean(governismo, na.rm=TRUE),
            freq = n()) %>%
  dplyr::group_by(legislator_id, legislator_name) %>%
  dplyr::summarise(governismo = round(mean(governismo, na.rm=TRUE),3),
            soma = sum(soma, na.rm=TRUE),
            nao = sum(nao), .groups = 'drop') %>%
           filter(!is.na(governismo))



deputados_atuais = CamaraBR::listaDeputadosAtivos()


dat = left_join(deputados_atuais, governismo_legislator %>% select(legislator_id, governismo, soma), by = "legislator_id") %>% mutate(Gov80 = ifelse(governismo > .80, 1, 0)) %>% count(Gov80) 



dat$dup = duplicated(dat$legislator_id)

dat %>% filter(dup==FALSE) %>% mutate(Gov80 = ifelse(governismo > .80, 1, 0)) %>% count(Gov80)