如何将mongoDB文档转换为python？

Question

我有一个mongoDB文档，我想将其转换为pandas数据框

 db.dataset2.insert(
{ 
 "user_id"    : "user_3",
 "order_id"   : "order_3",
 "order_lat " : -73.9557413,                       ## Order location
 "order_long" : 40.7720266,
 "order_time" : datetime.utcnow(),

 "dish" : [
      {
        "dish_id"          : "005" ,
        "dish_name"        : "Sandwitch",
        "dish_substitute"  : "Yes",
        "substitute_name"  : "Null",
        "dish_type"        : "Veg",               ## Binary response (Veg or Non-Veg)  
        "dish_price"       : 50,
        "dish_quantity"    : 1,
        "ratings"          : 3,
        "reviews"          : "blah blah blah",
        "home_chef_name"   : "ghyty",
        "expert_chef_name" : "abc" ,
        "coupon_applied"   : "Yes",               ## Binary response (Yes or No)
        "coupon_type"      : "Rs 20 off"
      },
      {
        "dish_id"          : "006" ,
        "dish_name"        : "Chicken Hundi",
        "dish_substitute"  : "No",
        "substitute_name"  : "Null",
        "dish_type"        : "Non-Veg",
        "dish_price"       : 125,
        "dish_quantity"    : 1,
        "ratings"          : 3,
        "reviews"          : "blah blah blah",
        "home_chef_name"   : "rtyu",
        "expert_chef_name" : "vbghy" ,
        "coupon_applied"   : "No",
        "coupon_type"      : "Null"

      }
   ],

}）

当我执行以下操作时

df = pd.DataFrame(list(db.dataset2.find()))

它给了我以下输出

           _id  \
  0  566148e3691db01e0cac9d82   
  1  56615926691db01e0cac9d83   
  2  56615c64691db01e0cac9d84   

                             dish                     order_id  order_lat   
0  [{u'dish_substitute': u'Yes', u'home_chef_name...  order_1  -73.955741   
1  [{u'dish_substitute': u'Yes', u'home_chef_name...  order_2  -73.955741   
2  [{u'dish_substitute': u'Yes', u'home_chef_name...  order_3  -73.955741   

    order_long       order_time          user_id  
0   40.772027   2015-12-04 08:03:47.658  user_1  
1   40.772027   2015-12-04 09:13:10.642  user_2  
2   40.772027   2015-12-04 09:27:00.497  user_3 

菜是json数组。 当我将其转换为数据框时，它会添加盘列，并将所有内容放在该列下。 我想将其转换为数据框以进行数据探索。 怎么做？ 我希望将其转换为以下格式。

      _id                   order_id   order_lat   order_long  
0  566148e3691db01e0cac9d82  order_1  -73.955741   40.772027   
1  566148e3691db01e0cac9d82  order_1  -73.955741   40.772027   

  order_time              user_id  coupon_applied coupon_type  dish_id  
0 2015-12-04 08:03:47.658  user_1            Yes   Rs 20 off     001   
1 2015-12-04 08:03:47.658  user_1             No        Null     001   

     dish_name      dish_price  dish_quantity dish_substitute dish_type  
0   Chicken Biryani     120              1             Yes     Non-Veg   
1   Paneer Biryani      100              1             Yes       Veg   

 expert_chef_name home_chef_name ratings  reviews          substitute_name  
0        abc            xyx        4   blah blah blah            Rice  
1        abc            abc        3   blah blah blah         Paratha 

请帮助..在此先感谢:)

Answer 1

你可以创建一个临时的DataFrame从记录df.dish并加入它回到原来的df 。

像这样：

df = pd.DataFrame(list(db.dataset2.find()))
tf = pd.DataFrame.from_records(df.dish)
df = df.join(tf)