单季度报表计算

说白了还是关于pandas包包的操练。

备忘:

1. 列的drop(基本)；

2. dataframe的merge；

###这个计算方法可以得到最具时效性的单季度报表
import pandas as pd
######################
def get_quarterly_financials(data_points,quarters,max_info_date,order_book_ids):
    datasets = pd.DataFrame()
    databulk = pd.DataFrame()
    
    #Begin: financial pit表的字典
    fields_table_map = {}
    for table_name,table in pit_financials.__dict__.items():
        if table_name.startswith('_'):
            continue
        for f in table.__dict__:
            if not f.startswith('_'):
                fields_table_map[f] = table_name
    financial_dict = fields_table_map    
    #End
    
    
    
    for data_point in data_points:
        if financial_dict.get(data_point) != 'balance_sheet' :  #非资产负债表的数据点的单季计算
            for quarter in quarters:
                for order_book_id in order_book_ids:
                    dataset = pd.DataFrame()
                    dataset = get_pit_financials(data_point,quarter,max_info_date,order_book_id)
                    if not dataset.empty is True:
                        dataset.pop('if_adjusted')
                        if int(quarter[-1]) - 1 == 0:  #是第一季度的时候就直接取最新的
                            dataset =  dataset.sort_values('info_date',ascending = False).head(1)
                            dataset.pop('info_date')
                            
                            datasets = datasets.append(dataset)
                            dataset = pd.DataFrame()
                        else: 
                            pre_quarter = quarter[:4] + 'q' + str(int(quarter[-1]) - 1)
                            pre_dataset = get_pit_financials(data_point,pre_quarter,max_info_date,order_book_id)
                            if not pre_dataset.empty is True:
                                pre_dataset.pop('if_adjusted')
                                pre_dataset = pre_dataset.sort_values('info_date',ascending = False).head(1)
                                dataset =  dataset.sort_values('info_date',ascending = False).head(1)
                                pre_dataset.pop('info_date')
                                dataset.pop('info_date')
                                
                                #需要判断下两个dp具体的值的情况
                                dataset[data_point] = dataset[data_point].values[0] - pre_dataset[data_point].values[0]
                                datasets = datasets.append(dataset)
                                dataset = pd.DataFrame()                  
                    else:
                        continue
        else:                                             #资产负债表字段的单季返回
            for quarter in quarters:  
                for order_book_id in order_book_ids:
                    dataset = get_pit_financials(data_point,quarter,max_info_date,order_book_id)
                    if not dataset.empty is True:
                        dataset = dataset.sort_values('info_date',ascending=False).head(1)
                        dataset.pop('info_date')
                        dataset.pop('if_adjusted')
                        datasets = datasets.append(dataset)
                        dataset=pd.DataFrame()
                
        if not databulk.empty is True:
            databulk = databulk.merge(datasets, how='outer', left_index=True, right_index=True)
            datasets = pd.DataFrame()
        else:
            databulk = databulk.append(datasets)
            datasets = pd.DataFrame()
    return databulk

#################
get_quarterly_financials(data_points=['operating_revenue','net_profit','total_assets'],quarters=['2018q1','2018q2'], max_info_date=None,\
                         order_book_ids=['000001.XSHE','000048.XSHE'])