Data processing before sending result back in MongoDB - mongodb

Question

We have a requirement to rank our records based on some defined algorithm. We have 4 fields in our MongoDB like following;
{
"rating" : 3.5
"review" : 4
"revenue" : 100
"used" : 3.9
},
{
"rating" : 1.5
"review" : 2
"revenue" : 10
"used" : 2.1
}
While querying the data, we will send % as weightage for our calculation. So assume we are sending 30% for rating, 30% for review and 20% each for revenue and uthe sed.
Now we need to score each record based on following calculation.
Score per column = (Existing Value - Average(Column) / StandardDeviation) * %weightage
for rating = (3.5 - 2.5) /1 * 30% = .03
So we need to count score for each column (or field) and than total of all 4 field will give a score to each record.
Is it possible to do such calculation with any MongoDB inbuilt function ?
Thanks in advance

Answer 1

Mongo has default operators for finding standard deviation ($stdDevPop) and for finding the average ($avg) and obviously for subtraction, multiplication, and division as well.
So, using all these operators you can definitely write an aggregation for what you require.
I've done for rating below, you can use the same logic for the other fields.
Also, replace 0.3 with your %weightage.
db.collection.aggregate([
{
$match: {
rating: {
$ne: null
}
}
},
{
$group: {
_id: null,
ratings: {
$push: "$rating"
},
avg_rating: {
$avg: "$rating"
},
std_deviation_rating: {
$stdDevPop: "$rating"
}
}
},
{
$project: {
ratings: {
$map: {
input: "$ratings",
as: "rating",
in: {
$multiply: [
{
$divide: [
{
$subtract: [
"$$rating",
"$avg_rating"
]
},
"$std_deviation_rating"
]
},
0.3
]
}
}
}
}
}
])