Преобразование r-кода в python pandas.code

Question

Я пытаюсь преобразовать этот код r в код панд.

Dependents.Numeric = recode(start.DF$Dependents, '3+'='3') %>% as.character() %>% as.numeric()
start.DF = start.DF %>% mutate(HouseholdSize = ifelse(is.na(Dependents.Numeric) | is.na(Married), NA,
                                                      ifelse(CoapplicantIncome > 0 | Married == "Yes",
                                                             Dependents.Numeric + 2,
                                                             Dependents.Numeric +1)))
start.DF = start.DF %>% mutate(IncomePC = TotalIncome/HouseholdSize)

Я хотел найти ответ на вопрос и нашел его, но он не в python и записан в r.Поэтому я был бы очень признателен, если бы кто-нибудь смог мне помочь.Спасибо

Answer 1

R код:

start.DF <- data.frame(
  Dependents = c('2', '3+', '1'),
  Married = c('yes', NA, 'no'),
  CoapplicantIncome = c(45, 0, 75),
  TotalIncome = c(100, 67, 80)
)
library(dplyr)
Dependents.Numeric = recode(start.DF$Dependents, '3+'='3') %>% as.character() %>% as.numeric()
start.DF = start.DF %>% mutate(HouseholdSize = ifelse(is.na(Dependents.Numeric) | is.na(Married), NA,
                                                      ifelse(CoapplicantIncome > 0 | Married == "Yes",
                                                             Dependents.Numeric + 2,
                                                             Dependents.Numeric +1)))
start.DF = start.DF %>% mutate(IncomePC = TotalIncome/HouseholdSize)
start.DF

  Dependents Married CoapplicantIncome TotalIncome HouseholdSize IncomePC
1          2     yes                45         100             4 25.00000
2         3+    <NA>                 0          67            NA       NA
3          1      no                75          80             3 26.66667

Код Python:

import pandas as pd
import numpy as np
start_df = pd.DataFrame.from_dict({
    'Dependents': ('2', '3+', '1'),
    'Married': ('yes', np.nan, 'no'),
    'CoapplicantIncome': (45, 0, 75),
    'TotalIncome': (100, 67, 80)
})

Dependents_Numeric = start_df['Dependents'].str.replace('3\+', '3').astype(float)
start_df['HouseholdSize'] = np.nan
cond1 = pd.isna(Dependents_Numeric) | pd.isna(start_df['Married'])
cond2 = (start_df['CoapplicantIncome'] > 0) | (start_df['Married'] == 'yes')
start_df.loc[~cond1 & cond2, 'HouseholdSize'] = Dependents_Numeric[~cond1 & cond2] + 2
start_df.loc[~cond1 & ~cond2, 'HouseholdSize'] = Dependents_Numeric[~cond1 & ~cond2] + 1
start_df['IncomePC'] = start_df['TotalIncome'] / start_df['HouseholdSize']
print(start_df)

   CoapplicantIncome Dependents Married  TotalIncome  HouseholdSize   IncomePC
0                 45          2     yes          100            4.0  25.000000
1                  0         3+     NaN           67            NaN        NaN
2                 75          1      no           80            3.0  26.666667