Question

我正在开发一个Django应用程序，它可以帮助分析来自不同位置的发射器的无线电信号传播。原始数据从MySQL数据库获得，该数据库记录报告者接收的每个信号及其来源和时间戳。

想法是在特定时间间隔/块期间检查接收器相对于目标发射器的状态。过滤后的结果需要作为JSON文件返回。

# All records within the time frame in question
qs_spots = Spot.objects.filter(unix_time__range = (starttime, endtime))
qs_spots = qs_spots_all.values()

# the focus is on a particular transmitter tx
tx_dict = {}
tx_dict['tx_callsign'] = tx

interval_range = 120

# list of all unique receivers - the field is called 'reporter' in the model
rx_list = qs_spots.values_list('reporter', flat=True).distinct()
tx_dict['rx_total'] = len(rx_list)
tx_dict['receivers'] = []

# goes through each reporter in the list
for rx in rx_list:
    rx_dict = {}
    rx_dict['rx_callsign'] = rx

    rx_spots = qs_spots.filter(reporter=rx)

    rx_dict['time_state'] = []

    current_time = starttime
    while current_time <= endtime:
        next_time = current_time + interval_range # increment next time interval
        time_dict = {}
        time_dict['datetime'] = current_time

        rx_time = rx_spots.filter(unix_time__range = (current_time, next_time))
        # first check if any records exist for the particular interval
        if rx_time.exists():                        
            # then check the transmitter status
            if rx_time.filter(transmitter=tx).exists():
                time_dict['status'] = 0 # reporter is receiving from target transmitter
            else:
                time_dict['status'] = 1 # reporter is not receiving from target transmitter, but is receiving from other transmitters
        else:
            # if no records exists for the particular interval
            # i.e reporter is not receiving at all during this interval
            time_dict['status'] = 2

        current_time = next_time # set to next interval

        rx_dict['time_state'].append(time_dict.copy())  

    tx_dict['receivers'].append(rx_dict.copy())

response = json.dumps(tx_dict, cls=DjangoJSONEncoder)
return HttpResponse(response, content_type='application/json')

我自己测试了外环并且它工作正常，虽然很慢 - 在30秒到2分钟之间返回50-100个接收器的JSON列表。当我添加内部循环以检查目标变送器的状态时，问题就来了。即使在数小时后，结果也永远不会恢复。

我已尝试过各种时间范围，但我的“默认”是在1小时窗口内以约120秒（2分钟）的间隔约50个接收器。 50 *（60/2）=总共1500个循环。对于计算机而言，听起来并不是很重要，但有些东西导致它被锁定。

编辑：添加了我的models.py（如果有帮助）

class Spot(models.Model):
    spot_id = models.IntegerField(db_column='Spot_ID', primary_key=True)  # Field name made lowercase.
    unix_time = models.IntegerField(db_column='UNIX_time', blank=True, null=True)  # Field name made lowercase.
    reporter = models.CharField(db_column='Reporter', max_length=15, blank=True, null=True)  # Field name made lowercase.
    transmitter = models.CharField(db_column='Transmitter', max_length=15, blank=True, null=True)  # Field name made lowercase.

class Meta:
    managed = False
    db_table = 'SPOT'

编辑2： 尝试将查询集转换为字典并限制字段

qs_spots = qs_spots.values('unix_time', 'reporter', 'transmitter')

rx_spots = qs_spots.filter(reporter=rx).values('unix_time', 'reporter', 'transmitter')

rx_time = rx_spots.filter(unix_time__range = (current_time, next_time)).values('transmitter')

外部循环似乎运行得更快，但内部循环仍然被锁定。

预期的JSON响应示例

{
  "tx_callsign": "EA4DUT",
  "rx_total": 3,
  "receivers": [
    {
      "rx_callsign": "G1RWT"
      "time_state": [
        { "datetime": 1443963600, "status": 0 },
        { "datetime": 1443963720, "status": 1 },
        { "datetime": 1443963840, "status": 2 }
        ]
    },
    {
      "rx_callsign": "PI4THT"
      "time_state": [
        { "datetime": 1443963600, "status": 1 },
        { "datetime": 1443963720, "status": 0 },
        { "datetime": 1443963840, "status": 2 }
        ]
    },
    {
      "rx_callsign": "EI6KD"
      "time_state": [
        { "datetime": 1443963600, "status": 2 },
        { "datetime": 1443963720, "status": 0 },
        { "datetime": 1443963840, "status": 0 }
        ]
    },
    ]
}

在嵌套循环中优化Django查询

0 个答案: