Scheduling Filter on Database with Time Interval

Question

Let's you have a science experiment called Weather and it records general information about the weather, one of which is recording the humidity every minute. After say 10 hours, you'll have 600 database values stored.

Weather.objects.filter().exclude(time__lt = commence).exclude(time__gt = cease)

This would create a filter for the Weather table, between the commence and cease times.

Let's say the time criteria is equal to -4 hours from present, therefore it would return the last 4 hours worth of data which is equal to 300 values. How would I ask for an interval of 10? Meaning, from these 300 values, how would I receive the 10th one of each, so 10, 20, 30, 40 ... 280, 290, 300... which would mean 30 values are being returned from the filter.

EDIT:

Would it be possible to ask for an interval of say 10 seconds? As opposed to each individual 10th value..

Answer 1

SQL doesn't have a way to ask for every tenth, so ORMs won't typically give you a way to express that need. You could do that work in Python:

samples = list(Weather.objects.filter().exclude(time__lt=commence).exclude(time__gt=cease))
for s10 in samples[::10]:
    # do something with every tenth...

This will pull all the samples from the database in one query, then only use 10% of them. You could also pull only the ones you need, one to a query:

sample_count = 600 # Somehow determine how many samples are in your time window.
samples = Weather.objects.filter().exclude(time__lt=commence).exclude(time__gt=cease)
for s_num in range(0, sample_count, 10):
    s = samples[s_num]
    # Now you have one sample in s.

To elaborate: range takes three arguments, start, stop, and step. For example:

>>> range(0, 100, 10)
[0, 10, 20, 30, 40, 50, 60, 70, 80, 90]

In the second example, the range expression is used to create the index values of the samples we want. Then samples[s_num] evaluates the QuerySet to get just that one value from the database.