American Community Survey: वार्षिक परिवर्तन

Python में US Census डेटा का विश्लेषण

Lee Hachadoorian

Asst. Professor of Instruction, Temple University

Census इतिहास: काउंट्स और सैंपल

मुख्य जनसांख्यिकीय विशेषताओं की पूर्ण गणना:

दशकीय Census 1790 - 2010+

व्यापक सामाजिक और आर्थिक विशेषताओं का सैंपल:

दशकीय Census "Long Form" (SF3) 1970 - 2000, ~15% परिवार
वार्षिक American Community Survey 2005+, ~1% परिवार

B25045 - Tenure by Vehicles Available by Age

Variable   |  Label
-----------|--------------------------------------
B25045001  |  Total
B25045002  |    Owner Occupied
B25045003  |      No Vehicle Available
B25045004  |        Householder 15 to 34 Years
B25045005  |        Householder 35 to 64 Years
B25045006  |        Householder 65 Years and Over
B25045007  |      1 or More Vehicles Available
B25045008  |        Householder 15 to 34 Years
B25045009  |        Householder 35 to 64 Years
B25045010  |        Householder 65 Years and Over
B25045011  |    Renter Occupied
B25045012  |      No Vehicle Available
B25045013  |        Householder 15 to 34 Years
B25045014  |        Householder 35 to 64 Years
B25045015  |        Householder 65 Years and Over
B25045016  |      1 or More Vehicles Available
B25045017  |        Householder 15 to 34 Years
B25045018  |        Householder 35 to 64 Years
B25045019  |        Householder 65 Years and Over

ACS विस्तृत टेबल अनुरोध - सेटअप

import requests
import pandas as pd

HOST, dataset = "https://api.census.gov/data", "acs/acs1"

get_vars = ["B25045_" + str(i + 1).zfill(3) + "E" for i in range(19)]
get_vars = ["NAME"] + get_vars

print(get_vars)

['NAME', 'B25045_001E', 'B25045_002E', 'B25045_003E', 'B25045_004E', 
'B25045_005E', 'B25045_006E', 'B25045_007E', 'B25045_008E', 'B25045_009E', 
'B25045_010E', 'B25045_011E', 'B25045_012E', 'B25045_013E', 'B25045_014E', 
'B25045_015E', 'B25045_016E', 'B25045_017E', 'B25045_018E', 'B25045_019E']

ACS विस्तृत टेबल अनुरोध - सेटअप

import requests
import pandas as pd

HOST, dataset = "https://api.census.gov/data", "acs/acs1"

get_vars = ["B25045_" + str(i + 1).zfill(3) + "E" for i in range(19)]
get_vars = ["NAME"] + get_vars

# print(get_vars)

predicates = {}
predicates["get"] = ",".join(get_vars)
predicates["for"] = "us:*"

कई वर्षों से वही वेरिएबल मंगाना

# DataFrame संग्राहक आरंभ करें
dfs = []

for year in range(2011, 2018):

    base_url = "/".join([HOST, str(year), dataset])
    r = requests.get(base_url, params=predicates)

    df = pd.DataFrame(columns=r.json()[0], data=r.json()[1:])

    # वर्ष मान रखने का कॉलम जोड़ें
    df["year"] = year

    dfs.append(df)


# संग्राहक में सभी DataFrames जोड़ें
us = pd.concat(dfs, ignore_index=True)

कई वर्षों से वही वेरिएबल मंगाना

print(us.head())

            NAME B25045_001E B25045_002E  ... B25045_019E us  year
0  United States   114991725    74264435  ...     3232812  1  2011
1  United States   115969540    74119256  ...     3447172  1  2012
2  United States   116291033    73843861  ...     3662322  1  2013
3  United States   117259427    73991995  ...     3847400  1  2014
4  United States   118208250    74506512  ...     4044430  1  2015

[5 rows x 22 columns]

चलो थोड़ा डेटा लें!

Python में US Census डेटा का विश्लेषण