Assignment 4

[wesbasinger]

Did you get stuck on Assignment 4 convert_housing_data_to_quarters() as well?

housing = pd.read_csv("City_Zhvi_AllHomes.csv")
housing.drop(housing.ix[:, '1996-04':'1999-12'], axis=1, inplace=True)
housing.drop(housing.ix[:, '2017-01':'2017-06'], axis=1, inplace=True)
housing.iloc[:,6:] = housing.iloc[:,6:].apply(pd.to_datetime, errors='coerce')
target = housing.iloc[:,6:].apply(pd.to_datetime, errors='coerce')
target = target.resample('Q', axis=1).mean()

I can't resample correctly...I wonder why...

[goddard-guryon]

I reached this far:

def convert_housing_data_to_quarters():
    housing_df = pd.read_csv('City_Zhvi_AllHomes.csv')
    State, Region = housing_df['State'], housing_df['RegionName']
    StateName = []
    for entry in State:
        StateName.append(states[entry])
    housing_df['StateName'] = StateName
    housing_df = housing_df.set_index(['StateName', 'RegionName'])
    housing_df = housing_df[housing_df.columns[-200:]]
    recess_df = pd.ExcelFile('gdplev.xls').parse(skiprows=7).loc[212:].reset_index()
    recess_df = recess_df[['Unnamed: 4', 'Unnamed: 5']]
    recess_df = recess_df.rename(columns={'Unnamed: 4': 'Quarter', 'Unnamed: 5': 'GDP'})
    recess_df['GDP'] = pd.to_numeric(recess_df['GDP'])
    for quarter in recess_df['Quarter']:
        count = 1
        housing_df[quarter] = housing_df[count:count+2].mean(axis=1)
        count += 3
    housing_df['2016q3'] = housing_df[-2:-1].mean(axis=1)
    housing_df = housing_df[housing_df.columns[-67:]]
    return housing_df

I did get the desired dimensions, but this gives a lot of NaN values in the DataFrame. Any clues on how to solve that?