Question

美好的一天！您能帮我查询吗？

我有一个表格“付款”：

import loompy as lp
import tensorflow as tf
from sklearn.model_selection import train_test_split

model_input_name = ""
input_size = 10000
batch_size = 32
epochs = 10

# Input functions for train, test and eval sets.
def train_input_fn():
    return _input_fn('TRAIN')

def test_input_fn():
    return _input_fn('TEST')

def eval_input_fn():
    return _input_fn('EVAL')

# General purpose input function
def _input_fn(mode = 'TRAIN'):

    """
        Arguments
            mode : 'TRAIN', 'TEST', 'EVAL'
    """

    # A generator to yield data and labels from the given FILE,
    # based on the indices assigned to the "indices" variable.
    # If you change the labels, remember to update the from_generator()
    # parameters below, to reflect their datatype.
    def gen():
        with lp.connect(FILE, 'r') as ds:
            if ae:
                for i in indices:
                    yield {model_input_name: ds[:, i]}, ds[:, i]
            else:
                for i in indices:
                    yield {model_input_name: ds[:, i]}, ds.ca.x_CellType[i]

    # Get the indices for train, test and eval sets
    train_idx, test_idx, eval_idx = train_test_set_idx_split(TRAIN_RT, TEST_RT, EVAL_RT)

    # Check condition and assign the respective set to the "indices" variable
    if mode == 'TRAIN':
        indices = train_idx
    elif mode == 'TEST':
        indices = test_idx
    elif mode == 'EVAL':
        indices = eval_idx
    else:
        print("Wrong mode choice: ", mode)
        exit(1)

    dataset = tf.data.Dataset.from_generator(gen, ({model_input_name: tf.int64}, tf.int64),
                                             output_shapes=({model_input_name: [input_size,]}, []))

    # Shuffle, batch, map, prefetch and repeat your dataset.
    # If you need to do some preprocessing on the data, create your function on
    # the cell above, and call it within a map() function.

    dataset = dataset.shuffle(buffer_size=batch_size*50)
    dataset = dataset.batch(batch_size)

    dataset = dataset.map(_reshape_labels)
    dataset = dataset.map(_int2float)

    # Map on whatever other functions you need
    dataset = dataset.map( ... )

    dataset = dataset.prefetch(2)
    dataset = dataset.repeat(epochs)

    iterator = dataset.make_one_shot_iterator()

    return iterator.get_next()


# Get train, test, eval indices for the given dataset
def train_test_set_idx_split(train_rt, test_rt, eval_rt):
    """ This function returns indices for the train, test and evaluation sets,
        given an input Dataset.
        Arguments:
            train_rt: ratio of the train dataset
            test_rt:  ratio of the test dataset
            eval_rt:  ratio of the evaluation dataset

        Returns:
            train_idx: indices (of the given dataset) for the train dataset
            test_idx:  indices (of the given dataset) for the test dataset
            evel_idx:  indices (of the given dataset) for the evaluation dataset

        Note:
            This function will work correctly as long as (test_rt == evel_rt) is True.
            If you need (test_rt != evel_rt), you need something more sophisticated.
    """

    with lp.connect(FILE, 'r') as ds:
        idx = np.array(range(0, ds.shape[1]))

    train_idx, test_idx = train_test_split(idx, train_size=train_rt, test_size=test_rt+eval_rt)
    test_idx, eval_idx = train_test_split(test_idx, train_size=0.5, test_size=0.5)

    return train_idx, test_idx, eval_idx

# Reshape labels as needed
def _reshape_labels(data, labels):
    return data, tf.reshape(labels, (-1,1))

任务是选择2012年1月1日至2012年3月31日的总付款金额，并根据用户曾经支付的金额将该金额按组划分。

组为“ 0-10”-如果总和为0 -10 $ “ 10或更多”-如果总和> 10 $。

我的代码：

payments
user_id     amount     payment_time    sale_type
1           20          31.01.2011      card
1           10          02.01.2012      cash
3           10          03.01.2012      card
4           15          05.02.2012      cash
...and so on

这是怎么了？

预期产量

SELECT * from (select IFnull(t.diapason,'total') as diapason, total_amount 
FROM 
    (SELECT p.user_id, p.amount as total_amount, CASE
        when amount<=10 then '0-10'
        when amount>10 then '10 and more' END AS diapason
           FROM (SELECT distinct payments.user_id, SUM(amount) AS amount 
                FROM payments inner JOIN (SELECT DISTINCT user_id 
                     FROM payments where payment_time between '2012-01-01' 
                     and '2012-01-30') a ON payments.user_id = a.user_id 
          GROUP BY payments.user_id) p) t  GROUP BY diapason WITH ROLLUP) as 
 t1  ORDER BY total_amount desc;

Answer 1

尝试此查询-

select case when p2.amount <=10 then '0-10'
        else '10 and more' end diapason
      ,p1.amount "total amount"
      ,p1.payment_by_card
      ,p1.cash
from (select user_id, sum(amount) amount, payment_by_card, cash
      from payments
      where payment_time between '2012-01-01' and '2012-01-30'
      group by user_id, payment_by_card, cash) p1
join (select user_id, sum(amount) amount
      from payments
      group by user_id) p2
on p1.user_id = p2.user_id

这是小提琴-http://www.sqlfiddle.com/#!9/22caaa/8

如何将计算结果按组划分？

1 个答案: