解决Jupyter Notebook中for循环输出DataFrame不够美观

Outline

用jupyter处理数据时，需要循环打印多个DataFrame，但打印出来的df看起来很难看；

想要的效果是比较规整美观的df展示，例如单独展示df时那样，是一个完整的table视图。

下面美化下jupyter notebook中for循环输出DataFrame

解决

解决前

for 循环打印DataFrame效果：

 看起来很不舒服

解决后

for 循环打印DataFrame效果：

 这下就舒服多了

解决方法

通过 from IPython.display import display 模块

代码中导入次模块，把print(df) 替换为 display(df)

from collections import Counter
import xml.etree.ElementTree as ET
from IPython.display import display
from pandas.io.json import json_normalize

# 需要解析的xml文件路径
path = '/data1/gaochao/refinitiv_sample/BGSBS02/BGSBS_03BDC.xml'  
tree = ET.parse(path)
root = tree.getroot()
iter_root = root.iter('BGSInformation')
ret = []

def get_xml_content(iter_root):
    for node in iter_root:
        dic = node.attrib
        text = node.text
        if len(dic) == 0 and text == '\n':
            pass
        else:
            dic['text'] = text
            ret.append({node.tag: dic})
        get_xml_content(node)
get_xml_content(iter_root)

def parse_data():
    # 获取标签名重复次数
    latest_ret = []
    conters = Counter([list(i.keys())[0] for i in ret])
    # 整合标签名只出现一次的
    for r_dict in ret:
        # 查看当前dict的key出现几次,如果只出现1次直接加入new_ret
        current_key = list(r_dict.keys())[0]
        repeat_num = conters[current_key]
        if repeat_num == 1:
            latest_ret.append(r_dict)
    # 整理标签名重复的字典
    for c in conters:
        # 查看当前dict的key出现几次,如果只出现1次直接加入new_ret
        current_key = c
        repeat_num = conters[current_key]
        if repeat_num != 1:
            # 出现几次,就去ret里拿几次
            tem_list = []
            for g in ret:
                if list(g.keys())[0] == current_key:
                    tem_list.append(g[current_key])
                else:
                    continue
            latest_ret.append({current_key: tem_list})
    return latest_ret

def get_df():
    # 生成DataFrame
    for p in parse_data():
        d_key = list(p.keys())[0]
        d_value = p[d_key]
        if type(d_value) is list:
            df = json_normalize(p, d_key)
            df_columns = df.columns.tolist()
            new_columns = [d_key + '_' + column for column in df_columns]
            df.set_axis(new_columns, axis='columns', inplace=True)
            display(df)
#             print(df)
        else:
            tem_dict = {}
            dict_value = p[d_key]
            for k in dict_value:
                tem_dict[d_key+'_'+k] = dict_value[k]
#             print(json_normalize(tem_dict))
            display(json_normalize(tem_dict))

get_df()