Foreign keys as QuerySet columns

Question

So I have 3 classes, a group class, a sensor class and a value class. The classes are roughly defined like this:

class SensorGroup(models.Model):
    name = models.CharField(max_length=100)
    logging_on = models.BooleanField(default=False)

class Sensor(models.Model):
    device_id = models.CharField(max_length=25, primary_key=True)
    name = models.CharField(max_length=25, blank=False)
    last_updated = models.DateTimeField(auto_now=True)
    sensor_group = models.ForeignKey(SensorGroup, related_name='sensors')

class Value(models.Model):
   sensor = models.ForeignKey(Sensor, related_name='values')
   time = models.DateTimeField(auto_now_add=True)
   value = models.FloatField(blank=False)

The values are added for every sensor at 20 second intervals. I want to turn it into a table like this using pandas.DataFrame.from_records() say for group 1:

| sensor_1 | sensor_2 | sensor_3 |    timestamp |
|----------+----------+----------+--------------|
|      1.5 |      2.0 |      1.0 | 12:33 2/4/17 |
|----------+----------+----------+--------------|
|      2.0 |      1.5 |      3.3 | 12:34 2/4/17 |
....

What I've tried so far:

qs = Sensors.all().values('name', 'values__value')
df = pd.DataFrame.from_records(list(qs))
print(df)

Yields

               name  values__value
0          sensor_1            0.0
1          sensor_1            1.0
2          sensor_1            2.0
3          sensor_2            0.0
4          sensor_2            1.0
5          sensor_2            2.0

Is there a way to make this query and convert it into a dataset?

Answer 1

Have a look at pandas pivot_table , if you extract the timestamp as well in the queryset, you could then do something like:

table = pd.pivot_table(df, values='values__value', index=['timestamp'], columns=['name']).reset_index()

to generate the table