Aggregate Functions (Group By)

Aggregate Functions are a continuation of Custom Select. It is recommended to have a understanding of custom selects before continuing.

Introduction

When working with custom selects, it is possible to use welds to create queries that result in group by statements. These are intended to be used with custom output structs just like custom selects.

Normal group by requirements apply with welds group bys. If you are using a group by, all columns must be either aggregated or in the group by.

Basic example

Lets start with a simple example. Counting.

#![allow(unused)]
fn main() {
// Struct for collecting group by output
#[derive(Debug, WeldsModel)]
pub struct PeopleCount {
    pub name: String,
    pub count: i32,
}
}

To count the people by their names, typically you would select the names grouping by the name column. Something like this in SQL

SELECT count(id), name FROM people GROUP BY name

This can be written in welds in a similar way

#![allow(unused)]
fn main() {
    let query = Player::all()
        .select(|p| p.name)
        .select_count(|p| p.id, "count")
        .group_by(|p| p.name);
    let data: Vec<PeopleCount> = query.run(&client).await?.collect_into()?;
}

All the aggregate functions in welds follow the pattern select_*( lambda, output_name)

You are required to include a name for your output data column because welds populates your struct by column name not index.