Groups aggregate functions using rxcpp?

Question

I'm trying to get a gist of RxCpp, the native cpp implementation of reactive extensions from Microsoft, to see if I could use it in a project, but I'm having trouble wrapping my head around the concepts.

If I had an observable templated with the following structure:

struct Person
{
    std::string name;
    std::string sex;
    int age;
}

How would I create another observable which contained, grouped by sex, the count of people, min age, max age and average age of all the events?

I've looked at examples, and I can't see how to get more than one aggregate at a time.

Kirk Shoop · Accepted Answer · 2016-01-05T19:37:36.897

Use group_by to partition by gender and then combine the min/max/average reducers to produce the desired output per gender.

Updated with count, output and additional comments

This works for me:

#include "rxcpp/rx.hpp"
using namespace rxcpp;
using namespace rxcpp::sources;
using namespace rxcpp::subjects;
using namespace rxcpp::util;

using namespace std;

struct Person
{
    string name;
    string gender;
    int age;
};

int main()
{
    subject<Person> person$;

    // group ages by gender
    auto agebygender$ = person$.
        get_observable().
        group_by(
            [](Person& p) { return p.gender;},
            [](Person& p) { return p.age;});

    // combine min max and average reductions.
    auto result$ = agebygender$.
        map([](grouped_observable<string, int> gp$){
            // the function passed to combine_latest 
            // will be called once all the source streams
            // (count, min, max, average) have produced a 
            // value. in this case, all the streams are reducers
            // that produce only one value when gp$ completes.
            // thus the function is only called once per gender 
            // with the final value of each stat.
            return gp$.
                count().
                combine_latest(
                    [=](int count, int min, int max, double average){
                        return make_tuple(gp$.get_key(), count, min, max, average);
                    },
                    gp$.min(),
                    gp$.max(),
                    gp$.map([](int age) -> double { return age;}).average());
        }).
        // this map() returns observable<observable<tuple<string, int, int, int, double>>>
        // the merge() returns observable<tuple<string, int, int, int, double>>
        // a grouped observable is 'hot' if it is not subscribed to immiediatly (in this case by merge)
        // then the values sent to it are lost.
        merge();

    // display results
    result$.
        subscribe(apply_to([](string gender, int count, int min, int max, double avg){
            cout << gender << ": count = " << count << ", range = [" << min << "-" << max << "], avg = " << avg << endl;
        }));

    //provide input data
    observable<>::from(
        Person{"Tom", "Male", 32},
        Person{"Tim", "Male", 12},
        Person{"Stel", "Other", 42},
        Person{"Flor", "Female", 24},
        Person{"Fran", "Female", 97}).
        subscribe(person$.get_subscriber());

    return 0;
}

with the resulting output

Female: count = 2, range = [24-97], avg = 60.5
Male: count = 2, range = [12-32], avg = 22
Other: count = 1, range = [42-42], avg = 42

Thanks, that looks great. A couple of questions: Why do you have to use [=] and mutable in the first argument to group_by? Can you add some additional comments in the section that combines min max and average? I don't quite understand what's going on there. Finally, obviously you know rxxcpp very well since you worked on the library, any tips on learning how to use it effectively? — bpeikes, Jan 05 '16 at 19:18
the [=] capture and mutable were artifacts of a previous example, I have removed them and added more comments. The majority of the concepts needed to use rxcpp are common to the other Rx implementations. This is great since the other resources can be reused. This page is how I exposed my self to these shared concepts: http://reactivex.io/learnrx/. I am happy to discuss rxcpp specific questions. other resources are: docs - http://reactive-extensions.github.io/RxCpp/ and chat - https://gitter.im/Reactive-Extensions/RxCpp — Kirk Shoop, Jan 05 '16 at 19:55

Groups aggregate functions using rxcpp?

1 Answers1