# Matrix API

## Get Matrix Result

<mark style="color:blue;">`GET`</mark> `https://penguin-stats.io/PenguinStats/api/v2/result/matrix`

This endpoint allows you to get matrix result.

#### Query Parameters

| Name                | Type    | Description                                                                                                                            |
| ------------------- | ------- | -------------------------------------------------------------------------------------------------------------------------------------- |
| stageFilter         | string  | Do filter on final result by stage. It should be a list of stageIds separated by commas.                                               |
| itemFilter          | string  | Do filter on final result by item. It should be a list of itemIds separated by commas.                                                 |
| server              | string  | Indicate which server you want to query. We support 4 servers now: "CN", "US", "JP" and "KR". Default value is "CN".                   |
| show\_closed\_zones | boolean | Whether show closed zones and stages or not. Default value is false.                                                                   |
| is\_personal        | boolean | Whether use personal drop data or global one. If the value is true, please make sure userID is set in cookies. Default value is false. |

{% tabs %}
{% tab title="200 Get result matrix successfully." %}

```javascript
{
  "matrix":[
    {
      "stageId":"main_01-07",
      "itemId":"30012",
      "quantity":164303,
      "times":131916,
      "start":1556676000000,
      "end":1589529600000
    },
    {
      "stageId":"main_01-07",
      "itemId":"30031",
      "quantity":7822,
      "times":131916,
      "start":1556676000000,
      "end":1589529600000
    },
    {
      "stageId":"sub_03-1-2",
      "itemId":"30032",
      "quantity":3162,
      "times":2079,
      "start":1556676000000,
      "end":1589529600000
    },
    {
      "stageId":"sub_03-1-2",
      "itemId":"randomMaterial_2",
      "quantity":20,
      "times":68,
      "start":1581105600000,
      "end":1582315200000
    }
  ]
}
```

{% endtab %}

{% tab title="500 Exceptions happen during the query." %}

```
```

{% endtab %}
{% endtabs %}

## What is Drop Matrix?

Suppose there is a matrix using stageId as row and itemId as column. Each element in the matrix represents the statistical data for that item in the stage. We will have a huge sparse matrix for all the items and stages, and for each server.

## Drop Matrix Element

### Model

```javascript
{
  stageId: string,
  itemId: string,
  quantity: number,
  times: number,
  start: number,
  end: number
}
```

### Properties

`stageId` and `itemId` are internal ids of stages and items. They can be found in [this repository](https://github.com/Kengxxiao/ArknightsGameData).

`start` is the left end (inclusive) of the interval. Time unit is millisecond.

`end` is the right end (exclusive) of the interval. Time unit is millisecond.

`quantity` means the number of dropped items during the interval.

`times` means the number of times this stage was played during the interval. **The value may be 0**, so please be careful when calculating the drop rate.

### Interval

Not all submitted drop samples will be used in the calculation of this matrix. As you can see above, each matrix element has its own interval. We design a special mechanism to decide what each interval should be.

Let's take stage `main_01-07` as an example. We have four time ranges A to D:

| Time Range | Meaning                                             | start         | end           |
| ---------- | --------------------------------------------------- | ------------- | ------------- |
| A          | Arknights CN released \~ Chapter 6 released         | 1556676000000 | 1577174400000 |
| B          | Chapter 6 released \~ 应急理智小样 event ended            | 1577174400000 | 1578340800000 |
| C          | 应急理智小样 event ended \~ 1st anniversary event started | 1578340800000 | 1589529600000 |
| D          | 1st anniversary event                               | 1589529600000 | 1590696000000 |

Consider the following two items:

* `30012`: Chinese name is "固源岩". It shows up in the drop list of `main_01-07` across all time ranges from A to D.
* `ap_supply_lt_010`: Chinese name is "应急理智小样". It only shows up in the drop list of `main_01-07` during time range B and D.

The calculation is assuming that the drop of `ap_supply_lt_010` is independent from other items in `main_01-07`. The `quantity` and `times` of `30012` can be summed up across time range A to D. So the summation interval of `30012` will be A to D, which means `[1556676000000, 1590696000000)`. Here we say all ranges for `30012` in `main_01-07`is **accumulatable**. (Maybe we should use the term "accumulable", however in the code and database we are using the former one, so we want to keep it as it is.)

For `ap_supply_lt_010`, we are not sure if the drop rate in time range D is different from  the one in B. Thus, we think that samples from B and D should not be summed up together.  We would like to say `ap_supply_lt_010` in `main_01-07` is non-accumulatable during time range B and D. Our matrix API is designed to reveal the latest drop rates, so the interval of `ap_supply_lt_010` in `main_01-07` in the matrix will be `[1589529600000, 1590696000000)`.

Although the data of `ap_supply_lt_010` in `main_01-07` is not showing in the matrix result, you can still use Advanced Query to get them.

The interval may be varies for different items in one stage. That is why we show it using `start` and `end` in the drop matrix element.