'Index' object has no attribute 'levels'

Question

I have the following command:

def convert_housing_data_to_quarters():
    import pandas as pd 
    housing = pd.read_csv('City_Zhvi_AllHomes.csv')
    housing = housing.drop(housing.columns[6:51],axis=1)
    times = housing[housing.columns[6:len(housing.columns)]]
    def quarters(col):
        if col.endswith(("01","02","03")):
            s = col[:4] + "q1"
        elif col.endswith(("04", "05", "06")):
            s = col[:4] + "q2"
        elif col.endswith(("07", "08", "09")):
            s = col[:4] + "q3"
        else:
            s = col[:4] + "q4"
        return s  
    times = times.groupby(quarters,axis=1).mean()
    df = pd.concat([times, housing[['State','RegionName']]], axis=1)
    arrays = [housing['State'],housing['RegionName']]
    index = pd.MultiIndex.from_arrays(arrays)
    index = index.sortlevel(level=0)
    df = df.reindex(index,level=0)
    return df 
convert_housing_data_to_quarters()

However, I keep getting the error message:

'Index' object has no attribute 'levels'

I am trying to create a hierarchical multi-index with 'State' at the top of the index (level=0) followed by 'RegionName' (level=1).

Would anybody be able to give me a helping hand as to where I am going wrong?

Answer 1

You cannot use reindex , because there is no MultiIndex . So use DataFrame.set_index by both columns before aggregate, so solution is possible simplify:

def convert_housing_data_to_quarters():
    import pandas as pd 
    housing = pd.read_csv('City_Zhvi_AllHomes.csv')
    housing = housing.drop(housing.columns[6:51],axis=1)
    times = housing[housing.columns[6:len(housing.columns)]]
    times = times.set_index(['State','RegionName']).sort_index()
    def quarters(col):
        if col.endswith(("01","02","03")):
            s = col[:4] + "q1"
        elif col.endswith(("04", "05", "06")):
            s = col[:4] + "q2"
        elif col.endswith(("07", "08", "09")):
            s = col[:4] + "q3"
        else:
            s = col[:4] + "q4"
        return s  
    return times.groupby(quarters,axis=1).mean()
convert_housing_data_to_quarters()

Also is possible convert datimes to quarters instead your function:

def convert_housing_data_to_quarters():
    housing = pd.read_csv('City_Zhvi_AllHomes.csv')
    housing = housing.drop(housing.columns[6:51],axis=1)
    times = housing.iloc[:, 6:len(housing.columns)]
    times = times.set_index(['State','RegionName']).sort_index()

    quarters = pd.to_datetime(times.columns, format='%Y-%m').to_period('Q')
    return  times.groupby(quarters,axis=1).mean()