熊猫循环并追加失败

Question

So I'm trying to send a datframe of VINs to the nhtsa open source api and get the information back. The below works with replacing df['vin'] with an actual vin. However when adding the loop to my function and trying to append, the result is that I get a blank dataframe back instead of the information for the 12 or so VINs. What am I doing wrong?

Code below:

import pandas as pd
import requests
    #develop the data
z = pd.DataFrame(columns = ["vin"], data = ['LHJLC79U58B001633','SZC84294845693987','LFGTCKPA665700387','L8YTCKPV49Y010001',
                                                 'LJ4TCBPV27Y010217','LFGTCKPM481006270','LFGTCKPM581004253','LTBPN8J00DC003107',
                                                 '1A9LPEER3FC596536','1A9LREAR5FC596814','1A9LKEER2GC596611','1A9L0EAH9C596099',
                                                 '22A000018'])

z['manufacturer'] = ['A','A','A','A','B','B','B','B','B','C','C','D','D']



def nhtsa(df):
    '''
    sends VIN to NHTSA for data call
    '''
    for i in df:
        url = 'https://vpic.nhtsa.dot.gov/api/vehicles/DecodeVINValuesBatch/'
        post_fields = {'format': 'json', 'data': df['vin']};
        r = requests.post(url, data=post_fields);
        x = r.json()
        f = pd.DataFrame(x['Results'])
        g = f[['VIN','Make','Manufacturer','ManufacturerId','ManufacturerType', 'Model','ModelYear', 'ABS','VehicleType', 'BodyClass','DisplacementCC','ErrorCode', 'SuggestedVIN']]
        csv = pd.DataFrame()
        csv = csv.append(g, ignore_index = True)
        return csv

nhtsa(z)

Answer 1

Ok. So, thanks to @G.Anderson, I figured this out. First, the empty data frame should be outside of the loop. Next, the i, as stated by G.Anderson was not getting called out. I used a .loc statement to call what i is. Below is the code.

def nhtsa(df):
    '''
    sends VIN to NHTSA for data call
    '''
    df1 = pd.DataFrame()
    for i in current_trial.index:
        vin = df.loc[i, 'VIN']
        url = 'https://vpic.nhtsa.dot.gov/api/vehicles/DecodeVINValuesBatch/'
        post_fields = {'format': 'json', 'data': vin};
        r = requests.post(url, data=post_fields);
        x = r.json()
        f = pd.DataFrame(x['Results'])
        g = f[['VIN','Make','Manufacturer','ManufacturerId','ManufacturerType', 'Model','ModelYear', 'ABS','VehicleType', 'BodyClass','DisplacementCC','ErrorCode', 'SuggestedVIN']]
        df1 = df1.append(g)
    return df1