How to make reports on the following JSON?

Question

MongoDB

Hello!
Finally came to the sweetest part of my project - the reports.
Friends plz tell me how to get reports based on the following data?

{
        "_id" : ObjectId("56e6b657753c6d890f614820"),
        "user" : ObjectId("56d9793edf03960106872e76"),
        "comment" : null,
        "status" : ObjectId("56bd55d111b7ed891118a65b"),
        "client" : null,
        "created" : ISODate("2016-03-14T13:02:15.030Z"),
        "ticket" : 0,
        "km" : 0,
        "sum" : 20,
        "service" : {
                "drunkDriver" : false,
                "person" : false,
                "nanny" : false,
                "courier" : false
        },
        "isCash" : false,
        "preorder" : false,
        "direction" : {
                "from" : ObjectId("56dccdd9c82df9d8048e0f6c"),
                "to" : ObjectId("56dccd94c82df9d8048e0f6a")
        },
        "moment" : {
                "closed" : ISODate("2016-03-14T13:15:00Z"),
                "end" : ISODate("2016-03-14T13:45:00Z"),
                "met" : ISODate("2016-03-14T13:00:00Z"),
                "inplace" : ISODate("2016-03-14T13:14:00Z"),
                "notification" : ISODate("2016-03-14T12:50:00Z"),
                "start" : null
        },
        "driver" : ObjectId("56e6a3feb770b5bc0b9e7593"),
        "passengers" : null,
        "__v" : 6
}

I still have virtual fields, don't know if they will work?
Here they are:
total - order amount (calculated based on other order data)
tripTime and waitingTime - waiting time and trip time.
Friends, please help me to make the following reports:
Total number of orders and amount for the user field
Total number of orders and amount for the driver field
Total amount and number of orders for the created field
I would like a couple of examples Otherwise I
can’t even imagine how these reports can be made :)

Answer 1

if we assume that the data value contains a list of your data
var data = [{
"_id" : ObjectId("56e6b657753c6d890f614820"),
"user" : ObjectId("56d9793edf03960106872e76"),
"status" : ObjectId("56bd55d111b7ed891118"),
"created " : ISODate("2016-03-14T13:02:15.030Z"),
"sum" : 20}, ........ ];
then you can use the reduce function

data.reduce(function (result, o) {
    var key = o.user;
    var counter = 1;
    if (!(key in result)) {
        result.arr.push(result[key] = { 
            user: o.user,
            couter: counter, 
            total: o.sum
        });
    } else {
        result[key].total += o.sum;
        counter++;
    }
    return result;
}, { arr: [] }).arr;

get the result for the user field, and the total amount and number of downloads
By driver, do the same, simply use driver as a key instead of user.
Working example on jsFiddle

Answer 2

Hey! Maybe I didn’t quite understand the idea, but since mapReduce examples have already been written, maybe you should look towards creating reports based on the Apache Solr search engine? As a result, you will get the flexibility of rebuilding reports for new requirements + near-realtime + scalability. The idea there is this: You load data into Solr and use search and analytical queries (facet & stats components) to do the necessary aggregations. Here is a video with an example of building analytical reports https://www.youtube.com/watch?v=JtbEDef_p9U You can build reports directly in Banana or Kibana, on which dashboards are made to visualize reports, see https://www.youtube .com/watch?v=cqV5lPM2VVE and https://www.youtube.com/watch?v=ddgtvv4W8EM. I would choose Banana as the dasboard builder as it is much more flexible and more responsive to Apache Solr. I also implemented analytics using Solr with the HBase layer. The bottom line is that if you need to change data that has already arrived, then it’s better to use key-value storage for updates, and then I have a livy-hbase indexer configured that updates the data in Solr in portions. Here is a link with more details: bigdata-intips.blogspot.com/2016/06/real-time-apac...