R/NSF.R

In abresler/govtrackR: Federal government tracking tools

# https://www.nsf.gov/crssprgm/reu/list_result.jsp;jsessionid=DFB0057418103DF5567873842A8A34BC?unitid=5049&d-8001259-s=1&d-8001259-o=1&showItems=15&d-8001259-p=3

.split_column <-
  function(data,
           id_column = "idAward",
           column = "emailInvestigator",
           clean = F) {
    if (!data %>% hasName(column)) {
      glue("Data missing {column}") %>% message()
      return(data)
    }
    is_date <- column %>% str_detect("date")

    df <-
      data %>%
      select(one_of(id_column, column)) %>%
      separate_rows(column, sep = "\\|") %>%
      mutate_at(column,
                str_squish) %>%
      distinct()

    if (is_date) {
      clean <- F
      df <- df %>% mutate_at(column, mdy)
    }

    if (clean) {
      df <-
        df %>% clean_entity_data(entity_column = column) %>% select(-column)
      names(df) <- names(df) %>% str_remove_all("Clean")
    }

    df <-
      df %>%
      filter(!is.na(!!sym(column))) %>%
      group_by_at(id_column) %>%
      summarise(!!sym(column) := !!sym(column) %>% unique() %>% sort() %>% str_c(collapse = " | ")) %>%
      ungroup()

    data <-
      data %>%
      select(-one_of(column)) %>%
      left_join(df, by = id_column) %>%
      select(one_of(names(data)), everything())

    data

  }

.dictionary_nsf_names <-
  function() {
    tibble(
      nameNSF = c(
        "AbstractNarration",
        "AwardAmount",
        "AwardEffectiveDate",
        "AwardExpirationDate",
        "AwardID",
        "AwardInstrument",
        "AwardTitle",
        "CityNameInstitution",
        "CodeFoaInformation",
        "CodeOrganization",
        "CodeProgramElement",
        "CodeProgramReference",
        "CountryNameInstitution",
        "nameDirectorateFunding",
        "nameDivisionFunding",
        "EmailAddressInvestigator",
        "FirstNameInvestigator",
        "LastNameInvestigator",
        "MaxAmdLetterDate",
        "MinAmdLetterDate",
        "NameFoaInformation",
        "NameInstitution",
        "PhoneNumberInstitution",
        "ProgramOfficer",
        "RoleCodeInvestigator",
        "StartDateInvestigator",
        "StateCodeInstitution",
        "StateNameInstitution",
        "StreetAddressInstitution",
        "TextProgramElement",
        "TextProgramReference",
        "ZipCodeInstitution",
        "AwardTotalIntnAmount",
        "EndDateInvestigator",
        "ARRAAmount",
        "AwardNumber",
        "Title",
        "NSFOrganization",
        "Program(s)",
        "StartDate",
        "LastAmendmentDate",
        "PrincipalInvestigator",
        "State",
        "Organization",
        "ProgramManager",
        "EndDate",
        "AwardedAmountToDate",
        "Co-PIName(s)",
        "PIEmailAddress",
        "OrganizationStreet",
        "OrganizationCity",
        "OrganizationState",
        "OrganizationZip",
        "OrganizationPhone",
        "NSFDirectorate",
        "ProgramElementCode(s)",
        "ProgramReferenceCode(s)",
        "Abstract",
        "DirectorateOrganization",
        "DivisionOrganization"
      ),
      nameActual = c(
        "descriptionAbstract",
        "amountAward",
        "dateAwardEffective",
        "dateAwardExpiration",
        "idAward",
        "typeAwardInstrument",
        "titleAward",
        "cityInstitution",
        "codeFOIA",
        "idOrganization",
        "codeProgramElement",
        "codeProgramReference",
        "countryInstitution",
        "nameDirectorateFunding",
        "nameDivisionFunding",
        "emailInvestigator",
        "nameFirstInvestigator",
        "nameLastInvestigator",
        "dateAmmendmentRecent",
        "dateAmmendmentInitial",
        "nameFOIA",
        "nameInstitution",
        "telephoneInstitution",
        "nameProgramOfficer",
        "roleInvestigator",
        "dateStart",
        "codeStateInstitution",
        "nameStateInstitution",
        "addressStreetInstitution",
        "textProgramElement",
        "referenceProgramElement",
        "zipcodeInstitution",
        "amountAwardTotalIntn",
        "dateEnd",
        "amountAmericanRecoveryAct",
        "idAward",
        "nameAward",
        "codeOrganization",
        "namePrograms",
        "dateStart",
        "dateLastAmmendment",
        "namePrincipalInvestigator",
        "codeState",
        "nameOrganization",
        "nameProgramManager",
        "dateEndDate",
        "amountAwardedToDate",
        "namesCoPrincipalInvestigators",
        "emailInvestigator",
        "addressStreetOrganization",
        "cityOrganization",
        "stateOrganization",
        "zipcodeOrganization",
        "telephoneOrganization",
        "codeNSFDirectorate",
        "codesProgramElement",
        "codesReferences",
        "descriptionAward",
        "nameDirectorateFunding",
        "nameDivisionFunding"
      )
    )
  }

.munge_nsf_names <-
  function(data) {
    dict_names <- .dictionary_nsf_names()
    nsf_names <-
      names(data)

    actual_names <-
      nsf_names %>%
      map_chr(function(name) {
        df_row <-
          dict_names %>% filter(nameNSF == name)
        if (nrow(df_row) == 0) {
          glue::glue("Missing {name}") %>% message()
          return(name)
        }

        df_row$nameActual
      })

    data %>%
      set_names(actual_names)
  }

.parse_nsf_xml <-
  function(f) {
    data <- read_xml(f) %>% xml_contents()
    nodeset <- data %>%
      xml2::xml_children()

    data <-
      seq_along(nodeset) %>%
      map_dfr(function(x) {
        node <- nodeset[[x]]
        name_nsf <- node %>% xml_name()

        children <- xml_children(x = node) %>% length()

        if (children <= 1)  {
          text <-
            xml_text(x = node)
          d <- tibble(x, item = name_nsf, value = text)
          return(d)
        }

        data <-
          seq_along(children) %>%
          map_dfr(function(z) {
            node_child <- xml_children(x = node)
            item <- node_child %>% xml_name()
            text <- xml_text(x = node_child)
            tibble(item = str_c(item, name_nsf), value = text)
          }) %>%
          mutate(x)

        data
      })

    data <- data %>%
      group_by(item) %>%
      filter(value != "") %>%
      summarise(value = str_c(value, collapse = " | ")) %>%
      ungroup()

    nsf_names <- data$item

    data <-
      data %>% spread(item, value) %>%
      select(one_of(nsf_names))

    data
  }

#' NSF Research Grant urls
#'
#' This function acquires urls
#' for all National Science Foundation Research Grants
#'
#' @return \code{tibble}
#' @export
#'
#' @examples
#' dictionary_nsf_award_urls()
dictionary_nsf_award_urls <-
  function() {
    page <- "https://nsf.gov/awardsearch/download.jsp" %>% read_html()
    years <-
      page %>% html_nodes(".downloadcontent strong") %>% html_text() %>% str_squish()
    urls <-
      page %>% html_nodes(".downloadcontent a") %>% html_attr("href") %>% str_c("https://nsf.gov/awardsearch/", .)
    current_year <- Sys.Date() %>% year()
    data <-
      tibble(years, urlNSFData = urls) %>%
      separate(years,
               into = c("yearAward", "sizeDataMB"),
               extra = "merge",
               sep = "\\-") %>%
      mutate_all(str_squish) %>%
      mutate(
        sizeDataMB = parse_number(sizeDataMB),
        yearAward = parse_number(yearAward),
        isCurrentYear = yearAward == current_year,
        isHistoric1976Prior = is.na(yearAward)
      ) %>%
      select(isCurrentYear, isHistoric1976Prior, everything())

    data
  }

.parse_nsf_url <-
  function(url = "https://nsf.gov/awardsearch/download?DownloadFileName=1960&All=true",
           return_message = T) {
    tmp <-
      tempfile()

    file <- curl_download(url, tmp)

    unz_files <- unzip(file, exdir = "xml")
    .parse_nsf_xml_safe <- possibly(.parse_nsf_xml, tibble())
    all_data <-
      unz_files %>%
      map_dfr(function(x) {
        x %>% message()
        .parse_nsf_xml_safe(f = x)
      })

    all_data <-
      all_data %>% .munge_nsf_names()

    all_data <-
      all_data %>%
      mutate_at(c("idAward", "idOrganization"),
                as.numeric)

    amt_names <- all_data %>% select(matches("amount")) %>% names()

    all_data <-
      all_data %>%
      mutate_at(amt_names,
                list(function(x) {
                  x %>% as.numeric() %>% currency(digits = 0)
                }))

    if (all_data %>% hasName("dateAwardEffective")) {
      all_data <-
        all_data %>%
        select(idAward, dateAwardEffective) %>%
        separate_rows("dateAwardEffective", sep = "\\|") %>%
        mutate(dateAwardEffective = dateAwardEffective %>% str_trim() %>% mdy()) %>%
        group_by(idAward) %>%
        summarise(dateAwardEffective = dateAwardEffective %>% str_c(collapse = " | ")) %>%
        ungroup() %>%
        left_join(all_data %>% select(-dateAwardEffective), by = "idAward") %>%
        select(one_of(names(all_data)), everything())

    }

    if (all_data %>% hasName("dateAwardExpiration")) {
      all_data <-
        all_data %>%
        select(idAward, dateAwardExpiration) %>%
        separate_rows("dateAwardExpiration", sep = "\\|") %>%
        mutate(dateAwardExpiration = dateAwardExpiration %>% str_trim() %>% mdy()) %>%
        group_by(idAward) %>%
        summarise(dateAwardExpiration = dateAwardExpiration %>% str_c(collapse = " | ")) %>%
        ungroup() %>%
        left_join(all_data %>% select(-dateAwardExpiration), by = "idAward") %>%
        select(one_of(names(all_data)), everything())

    }

    date_cols <-
      all_data %>% select(matches("date")) %>% select(-one_of(c(
        "dateAwardExpiration",
        "dateAwardEffective"
      ))) %>%
      names()

    if (length(date_cols) > 0) {
      all_data <- all_data %>%
        mutate_at(date_cols, list(mdy))
    }

    if (all_data %>% hasName("dateAwardExpiration")) {
      all_data <- all_data %>%
        mutate(dateAwardExpiration = ymd(dateAwardExpiration))
    }

    if (all_data %>% hasName("dateAwardEffective")) {
      all_data <- all_data %>%
        mutate(dateAwardEffective = ymd(dateAwardEffective))
    }

    df_people <-
      all_data %>%
      select(idAward, nameFirstInvestigator) %>%
      separate_rows(nameFirstInvestigator, sep =  "\\|") %>%
      mutate_if(is.character, str_squish) %>%
      mutate(idRow = 1:n()) %>%
      left_join(
        all_data %>%
          select(idAward, nameLastInvestigator) %>%
          separate_rows(nameLastInvestigator, sep =  "\\|") %>%
          mutate(idRow = 1:n()) %>%
          mutate_if(is.character, str_squish),
        by = c("idAward", "idRow")
      ) %>%
      select(-idRow) %>%
      mutate_if(is.character,
                list(function(x) {
                  ifelse(is.na(x), "", x)
                })) %>%
      unite(
        nameInvestigator,
        nameFirstInvestigator,
        nameLastInvestigator,
        sep  = " ",
        remove = F
      ) %>%
      mutate(nameInvestigator = nameInvestigator %>% str_squish()) %>%
      group_by(idAward) %>%
      summarise_all(list(function(x) {
        x %>% unique() %>% sort() %>% str_c(collapse = " | ") %>% str_to_upper()
      })) %>%
      ungroup() %>%
      mutate_at("nameInvestigator", list(function(x) {
        ifelse(x %in% c("", "DATA NOT AVAILABLE"), NA_character_, x)
      }))

    all_data <-
      all_data %>% select(-c(nameFirstInvestigator, nameLastInvestigator)) %>%
      left_join(df_people, by = "idAward") %>%
      select(idAward, nameInvestigator, everything())

    all_data <-
      all_data %>% .split_column(column = 'countryInstitution', clean = T)
    all_data <-
      all_data %>% .split_column(column = 'nameFOIA', clean = T)
    all_data <-
      all_data %>% .split_column(column = 'cityInstitution', clean = T)
    all_data <-
      all_data %>% .split_column(column = 'nameInstitution', clean = T)
    all_data <-
      all_data %>% .split_column(column = 'roleInvestigator', clean = T)
    all_data <-
      all_data %>% .split_column(column = "typeAwardInstrument", clean = T)
    all_data <-
      all_data %>% .split_column(column = "nameDirectorateFunding", clean = T)
    all_data <-
      all_data %>% .split_column(column = "nameDivisionFunding", clean = T)
    all_data <-
      all_data %>% .split_column(column = "nameStateInstitution", clean = T)
    all_data <-
      all_data %>% .split_column(column = "addressStreetInstitution", clean = T)
    all_data <-
      all_data %>% .split_column(column = "textProgramElement", clean = T)
    all_data <-
      all_data %>% .split_column(column = "referenceProgramElement", clean = T)
    all_data <-
      all_data %>% .split_column(column = "nameProgramOfficer", clean = T)
    all_data <-
      all_data %>% .split_column(column = "nameDirectorateFunding", clean = T)

    if (all_data %>% hasName("descriptionAbstract")) {
      all_data <-
        all_data %>%
        mutate_at(c("descriptionAbstract"),
                  list(function(x) {
                    x %>% stri_enc_toascii() %>% str_remove_all(" \032 ") %>% str_to_upper() %>% str_replace_all("<BR/>|<BR>", " ") %>% str_squish() %>% str_remove("ABSTRACT") %>% str_squish()
                  }))
    }

    if (all_data %>% hasName("titleAward")) {
      all_data <-
        all_data %>%
        mutate_at(c("titleAward"),
                  list(function(x) {
                    x %>% stri_enc_toascii() %>% str_remove_all(" \032 ") %>% str_to_upper() %>% str_replace_all("<BR/>|<BR>", " ") %>% str_squish() %>% str_remove("ABSTRACT") %>% str_squish()
                  }))

    }

    all_data <-
      all_data %>%
      mutate(urlNSFData = url)

    file %>% unlink()
    unlink("xml", recursive = T)
    unz_files %>% unlink()

    if (all_data %>% hasName("zipcodeInstitution")) {
      all_data <-
        all_data %>%
        mutate(zipcodeInstitution = zipcodeInstitution %>% substr(1, 5))
    }

    if (all_data %>% select(dateAmmendmentInitial, dateAmmendmentRecent) %>% ncol() == 2) {
      all_data <-
        all_data %>%
        mutate(countDaysAmmendment = pmin(
          0,
          (dateAmmendmentRecent - dateAmmendmentInitial) %>% as.numeric()
        ))
    }

    if (all_data %>% select(dateAwardEffective, dateAwardExpiration) %>% ncol() == 2) {
      all_data <- all_data %>%
        mutate(countDaysContract = pmin((
          dateAwardExpiration - dateAwardEffective
        ) %>% as.numeric()))
    }

    gc()

    if (all_data %>% select(matches("address|city|codeState|zip")) %>% ncol() == 4) {
      df_locations <-
        all_data %>%
        select(
          idAward,
          addressStreetInstitution,
          cityInstitution,
          codeStateInstitution,
          zipcodeInstitution
        ) %>%
        mutate_if(is.character, list(function(x) {
          ifelse(is.na(x), "", x)
        })) %>%
        unite(cityState, cityInstitution, codeStateInstitution, sep = " ") %>%
        unite(cityStateZip, cityState, zipcodeInstitution, sep = " ") %>%
        unite(locationInstitution,
              addressStreetInstitution,
              cityStateZip,
              sep = ", ") %>%
        mutate(locationInstitution = str_squish(locationInstitution))

      all_data <-
        all_data %>%
        left_join(df_locations, by = "idAward") %>%
        select(
          one_of(
            "idAward",
            "titleAward",
            "nameInstitution",
            "nameInvestigator",
            "locationInstitution"
          ),
          everything()
        )

      all_data <-
        all_data %>%
        mutate(
          locationInstitution = case_when(
            locationInstitution == ", ," ~ NA_character_,
            locationInstitution %>% substr(1, 2) == ", " ~ locationInstitution %>% substr(3, nchar(locationInstitution)),
            locationInstitution %>% substr(nchar(locationInstitution), nchar(locationInstitution)) == "," ~
              locationInstitution %>% substr(1, nchar(locationInstitution) - 1),
            TRUE ~ locationInstitution
          )
        )
    }

    all_data <-
      all_data %>%
      mutate(
        urlAward = glue(
          "https://www.nsf.gov/awardsearch/showAward?AWD_ID={idAward}"
        ) %>% as.character()
      )

    all_data
  }

#' NSF years grants
#'
#' Download and clean grants
#' from the National Science Foundationf
#' for specified periods
#'
#' @param years vector of numeric years
#' @param clean_institutions if \code{TRUE} cleans institution names
#' @param clean_parties if \code{TRUE} cleans party names
#' @param return_message if \code{TRUE} returns message
#'
#' @return \code{tibble}
#' @export
#'
#' @examples
nsf_years_grants <-
  function(years = 2018,
           clean_institutions = T,
           clean_parties = T,
           return_message = T
  ) {

    df_years <-
      dictionary_nsf_award_urls()

    df_years <- df_years %>%
      select(yearAward, urlNSFData)

    urls <- df_years %>% filter(yearAward %in% years) %>% pull(urlNSFData)
    .parse_nsf_url_safe <- possibly(.parse_nsf_url, tibble())

    all_data <-
      urls %>%
      map_dfr(function(url){
        .parse_nsf_url(url = url,return_message = return_message)
      })

    all_data <-
      all_data %>%
      mutate_if(is.character,
                list(function(x){
                  ifelse(x == "", NA_character_, x) %>% str_squish()
                }))

    amt_names <- all_data %>% select(matches("amount")) %>% names()

    all_data <-
      all_data %>%
      left_join(df_years, by = "urlNSFData")

    all_data <-
      all_data %>%
      mutate_at(amt_names,
                list(function(x) {
                  x %>% as.numeric() %>% currency(digits = 0)
                })) %>%
      select(yearAward, everything())


    if (all_data %>% hasName("textProgramElement")) {
      if (all_data %>% hasName("descriptionAbstract")) {
        all_data <-
          all_data %>%
          mutate(
            isSBIRSTTR = textProgramElement %>% str_detect(
              "SBIR|STTR|SMALL BUSINESS PHASE|SMALL BUSINESS INNOVATION RESEARCH"
            ),
            haSBIRSTTRMention = descriptionAbstract %>% str_detect(
              "SBIR|STTR|SMALL BUSINESS PHASE|SMALL BUSINESS INNOVATION RESEARCH|SMALL BUSINESS TECHNOLOGY TRANSFER"
            ) |
              titleAward %>% str_detect(
                "SBIR|STTR|SMALL BUSINESS PHASE|SMALL BUSINESS INNOVATION RESEARCH|SMALL BUSINESS TECHNOLOGY TRANSFER"
              ),
            isSBIRSTTR = case_when(is.na(isSBIRSTTR) &
                                     haSBIRSTTRMention ~ TRUE,
                                   TRUE ~ isSBIRSTTR)
          )
      } else {
        all_data <-
          all_data %>%
          mutate(
            isSBIRSTTR = textProgramElement %>% str_detect(
              "SBIR|STTR|SMALL BUSINESS PHASE|SMALL BUSINESS INNOVATION RESEARCH"
            ),
            haSBIRSTTRMention = titleAward %>% str_detect(
                "SBIR|STTR|SMALL BUSINESS PHASE|SMALL BUSINESS INNOVATION RESEARCH|SMALL BUSINESS TECHNOLOGY TRANSFER"
              ),
            isSBIRSTTR = case_when(is.na(isSBIRSTTR) &
                                     haSBIRSTTRMention ~ TRUE,
                                   TRUE ~ isSBIRSTTR)
          )
      }

    } else {
      if (all_data %>% hasName("descriptionAbstract")) {
        all_data <- all_data %>%
          mutate(
            haSBIRSTTRMention = descriptionAbstract %>% str_detect(
              "SBIR|STTR|SMALL BUSINESS PHASE|SMALL BUSINESS INNOVATION RESEARCH|SMALL BUSINESS TECHNOLOGY TRANSFER"
            ) |
              titleAward %>% str_detect(
                "SBIR|STTR|SMALL BUSINESS PHASE|SMALL BUSINESS INNOVATION RESEARCH|SMALL BUSINESS TECHNOLOGY TRANSFER"
              ),
            isSBIRSTTR = haSBIRSTTRMention
          )
      } else {
        all_data <-
          all_data %>%
          mutate(
            haSBIRSTTRMention =  titleAward %>% str_detect(
                "SBIR|STTR|SMALL BUSINESS PHASE|SMALL BUSINESS INNOVATION RESEARCH|SMALL BUSINESS TECHNOLOGY TRANSFER"
              ),
            isSBIRSTTR = haSBIRSTTRMention
          )
      }

    }


    all_data

  }

.nsf_grants <-
  function() {
    file_nos <- 1:10
    urls <- glue("https://asbcllc.com/r_packages/govtrackR/data/nsf/nsf_grants/nsf_grants_{file_nos}.rda")

    data <-
      urls %>%
      map_dfr(function(url){
        read_rda(file = url)
      })

    amt_names <- data %>% select(matches("amount")) %>% names()

    data <-
      data %>%
      mutate_at(amt_names,
                list(function(x){
                  currency(x, digits = 0)
                })) %>%
    mutate(nameAgencyParent = "NATIONAL SCIENCE FOUNDATION")

    name_cols <- data %>% select(matches("name")) %>% names()

    data <- data %>%
      mutate_at(name_cols, str_to_upper)

    data <- data %>%
      separate(
        nameDivisionFunding,
        extra = "merge",
        into = c("nameDivisionFunding", "slugDivisionFunding"),
        sep = "\\("
      ) %>%
      mutate(slugDivisionFunding = slugDivisionFunding %>% str_remove_all("\\)")) %>%
      mutate_if(is.character, str_squish)

    data <- data %>%
      mutate(nameAgencyParent = "NATIONAL SCIENCES FOUNDATION")

    data

  }

#' National Science Foundation Grants
#'
#' Cached version of all National Science Foundation
#' Grants since 1971
#'
#' @return
#' @export
#'
#' @examples
nsf_grants <-
  function(){
    .tt <- memoise::memoise(.nsf_grants)
    .tt()
  }




# terms -------------------------------------------------------------------


.parse_nfs_grant_term <-
  function(term = "Deep Learning") {
    cookies = c(
      'JSESSIONID' = 'D53656E0B6FFCCF35301DA4FC8BAC9D1'
    )
    slug <- glue('"{term}"') %>% str_to_upper() %>% URLencode()
    url_refer <-
      glue('https://nsf.gov/awardsearch/advancedSearchResult?PIId=&PIFirstName=&PILastName=&PIOrganization=&PIState=&PIZip=&PICountry=&ProgOrganization=&ProgEleCode=&BooleanElement=All&ProgRefCode=&BooleanRef=All&Program=&ProgOfficer=&Keyword={slug}&AwardNumberOperator=&AwardAmount=&AwardInstrument=&ActiveAwards=true&OriginalAwardDateOperator=&StartDateOperator=&ExpDateOperator=') %>% as.character()

    headers = c(
      `Connection` = 'close',
      `Upgrade-Insecure-Requests` = '1',
      `DNT` = '1',
      `User-Agent` = 'Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_0) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/78.0.3904.70 Safari/537.36',
      `Sec-Fetch-User` = '?1',
      `Accept` = 'text/html,application/xhtml+xml,application/xml;q=0.9,image/webp,image/apng,*/*;q=0.8,application/signed-exchange;v=b3',
      `Sec-Fetch-Site` = 'same-origin',
      `Sec-Fetch-Mode` = 'navigate',
      `Referer` = url_refer,
      `Accept-Encoding` = 'gzip, deflate, br',
      `Accept-Language` = 'en-US,en;q=0.9'
    )

    params = list(`exportType` = 'csv')
    url_refer %>% message()
    res <-
      httr::GET(
        url = 'https://nsf.gov/awardsearch/ExportResultServlet',
        httr::add_headers(.headers = headers),
        query = params,
        httr::set_cookies(.cookies = cookies)
      )

    data <-
      res %>%
      content() %>%
      as_tibble()

    data <-
      .munge_nsf_names(data = data) %>%
      .munge_data(clean_address = F) %>%
      mutate(termSearch = term ) %>%
      select(termSearch, everything())

    data

  }

.parse_nsf_award_url <-
  function(url = "https://www.nsf.gov/awardsearch/showAward?AWD_ID=1610953") {
    page <- read_html(x = url)
    items <-
      page %>% html_nodes(".tabletext2 strong") %>% html_text()
    values <-
      page %>% html_nodes(".tabletext2+ .tabletext2")

    seq_along(items) %>%
      map_dfr(function(x){
        item <- items[[x]]

        if (x == 1) {
          values[x] %>% html_nodes("a") %>% html_text()
        }

      })
  }


.munge_nsf <-
  function(data) {



  }