Web API Result for a column in R-CodePudding

I have one column of a data frame that consists of names for more than one thousands.

Name
Barack Obama
Xijin Ping
Bladimir Putin
John Smith
...

I want to create a new column for ethnicity scores that will be retrieved from API (https://www.name-prism.com/api). I have received API token from the website and an example code for Barack Obama (http://www.name-prism.com/api_token/nat/csv/[API_token]/Barack Obama). The web results show as follows.

European-SouthSlavs,0.0000 Muslim-Pakistanis-Bangladesh,0.0000 European-Italian-Italy,0.0000 European-Baltics,0.0000 African-SouthAfrican,0.0000 European-Italian-Romania,0.0031 Muslim-Nubian,0.0026 European-French,0.1359 EastAsian-Indochina-Thailand,0.0000 EastAsian-Indochina-Vietnam,0.0108 Jewish,0.0000 Muslim-Turkic-CentralAsian,0.0000 EastAsian-Indochina-Cambodia,0.0000 Nordic-Scandinavian-Denmark,0.0000 EastAsian-Indochina-Myanmar,0.0000 Nordic-Finland,0.0000 Muslim-Persian,0.0035 Nordic-Scandinavian-Sweden,0.0000 Muslim-Maghreb,0.0000 Greek,0.0000 Muslim-Pakistanis-Pakistan,0.0000 Hispanic-Portuguese,0.0003 European-Russian,0.0128 Muslim-ArabianPeninsula,0.0000 African-WestAfrican,0.0324 EastAsian-Japan,0.0000 European-German,0.0001 EastAsian-Chinese,0.0005 SouthAsian,0.0060 Hispanic-Spanish,0.0126 Nordic-Scandinavian-Norway,0.0000 Muslim-Turkic-Turkey,0.0000 Hispanic-Philippines,0.0001 CelticEnglish,0.0436 EastAsian-Malay-Malaysia,0.0041 EastAsian-South Korea,0.0000 African-EastAfrican,0.7259 European-EastEuropean,0.0000 EastAsian-Malay-Indonesia,0.0057

Below I managed to get a result from the case of Barack Obama, but, I am not sure for more than one thousand names.

result <- GET("http://www.name-prism.com/api_token/nat/csv/[API_token]/Barack Obama")

The desired outcomes will be as follows. Based on the list of names in the data frame, I want to add the ethnicity score from the web API as a new column,

    Name // Ethnicity Score
    Barack Obama // 0.781
    Xijin Ping // 0.812
    Bladimir Putin // 0.912
    John Smith // 0.777
    ...`

Thank you for your help in advance!

CodePudding user response：

With the specification you give this is what i can do.

Suppose you have a data.frame called df_names and the column df_names$name has the names you are to search the ethnicities. This will create a new df df_result with the names and the top ethnicities and values (for the cases that were found).

However i think the API may have some limit on the number of requests you can do.

I hope this can help you.


df_result <- purrr::map_dfr(df_names$name, function(name) {
    result <- GET(paste0("http://www.name-prism.com/api_token/nat/csv/",
               api_token,"/", URLencode(name)))
    if(http_error(result)){
        NULL
    }else{
        eth<- content(result, "text")
        eth<- do.call(rbind, strsplit(strsplit(eth, split = "(?<=\\d) ", perl=T)[[1]],","))
        #first three ethincities
        top_eth <- eth[order(as.numeric(eth[,2]), decreasing = T)[1:3],]
    
        c(name,as.vector(t(top_eth)))
    }
})