MongoDB query to compute percentage - mongodb

Question

I am new to MongoDB and kind of stuck at this query. Any help/guidance will be highly appreciated. I am not able to calculate the percentage in the desired way. There is something wrong with my pipeline where prerequisites of percentage are not computed correctly. Following I provide my unsuccessful attempt along with the desired output.
Single entry in the collection looks like below:
_id : ObjectId("602fb382f060fff5419fd0d1")
time : "2019/05/02 00:00:00"
station_id : 3544
station_name : "Underhill Ave &; Pacific St"
station_status : "In Service"
latitude : 40.6804836
longitude : -73.9646795
zipcode : 11238
borough : "Brooklyn"
neighbourhood : "Prospect Heights"
available_bikes : 5
available_docks : 21
The query I am trying to solve is:
Given a station_id (e.g., 522) and a num_hours (e.g., 3) passed as parameters:
- Consider only the measurements where the station_status = “In Service”.
- Consider only the measurements for that concrete
“station_id”.
- Compute the percentage of measurements with
available_bikes = 0 for each hour of the day (e.g., for the period
[8am, 9am) the percentage is 15.06% and for the period [9am, 10am)
the percentage is
27.32%).
- Sort the percentage results in decreasing order.
- Return the top “num_hours” documents.
The desired output is:
--- DOCUMENT 0 INFO ---
---------------------------------
hour : 19
percentage : 65.37
total_measurements : 283
zero_bikes_measurements : 185
---------------------------------
--- DOCUMENT 1 INFO ---
---------------------------------
hour : 21
percentage : 64.79
total_measurements : 284
zero_bikes_measurements : 184
---------------------------------
--- DOCUMENT 2 INFO ---
---------------------------------
hour : 00
percentage : 63.73
total_measurements : 284
zero_bikes_measurements : 181
My attempt is:
command_1 = {"$match": {"station_status": "In Service", "station_id": station_id, "available_bikes": 0}}
my_query.append(command_1)
command_2 = {"$group": {"_id": "null", "total_measurements": {"$sum": 1}}}
my_query.append(command_2)
command_3 = {"$project": {"_id": 0,
"station_id": 1,
"station_status": 1,
"hour": {"$substr": ["$time", 11, 2]},
"available_bikes": 1,
"total_measurements": {"$sum": 1}
}
}
my_query.append(command_3)
command_4 = {"$group": {"_id": "$hour", "zero_bikes_measurements": {"$sum": 1}}}
my_query.append(command_4)
command_5 = {"$project": {"percent": {
"$multiply": [{"$divide": ["$total_measurements", "$zero_bikes_measurements"]},
100]}}}
my_query.append(command_5)

Answer 1

I've taken a look at this and I'm going to offer some sincere advice:
Don't try and do this in an aggregate query. Just go back to basics and pull the numbers out using find()s and then calculate the numbers in python.
If you want to persist with an aggregate query, I will say that your match command filters on available_bikes equal to zero. You never have the total number of measurements, so you can never find the percentage. Also when you have done your first $group, your "lose" your projection so at that point in the pipeline you only have total_measurements and that's it (comment out the commands 3 to 5 to see what I mean).