More Configuration Example

[jura-b]

Could you provide more example of each type of boot option, please?

[jonathan-casarrubias]

@jura-b the mixin is actually quite robust, but at the same time complex...

Do you have a specific result in mind? I can help on that... adding all types of examples might also confuse you

[jura-b]

@jonathan-casarrubias
Suppose I have a model schema like this,

{ title: "Example 1", result: { score: 10 } }

I want the average of result.score. How does the config look like?

Thank you in advance!

[jonathan-casarrubias]

@jura-b the stats mixing results in series of data expected to be passed to any chart, said that.

I won't provide the functionality aside of creating series, for instance: If you only want to get the average from a couple of records.

What actually does is to create series of data based on time:

Lets say you have the following records

{ title: "Example 1", result: { score: 10 }, createdAt: '2017/07/15' }
{ title: "Example 2", result: { score: 15 }, createdAt: '2017/07/15' }
{ title: "Example 3", result: { score: 10 }, createdAt: '2017/07/16' }

*Note date should actually be ISODate (Recommended use of loopback-ds-timestamp-mixin)

When using the stats mixing it will return series which in this use case one of the datum would be 12.5 from the average in day 15.

To achieve that you would configure as follows:

"mixins": {
    "Stats": [
        {
            "method": "average", // This can be whatever you want
            "endpoint": "/average", // This also can be whatever you want
            "description": "A Description for your average method", // As well as this
            "type": "model", // We will calculate over this model, not from a related model
            "count": {
                "on": "createdAt", // This creates time based series
                "by": "result.score", // will count by result.score
                "avg": true // This will return the average from result.score (12.5 at Aug 15)
            }
        }
    ]
}

I hope this clarifies your question and also the mixing functionality.

Regards
Jon

[jura-b]

@jonathan-casarrubias Really appreciate your help!

[sitthinutk]

@jonathan-casarrubias
Hi Jon,
I have some relatable question with that of @jura-b , but a little bit more complex.
Suppose I have records with schema like this,

{ title: "Exam A",
  answer: [{_id: 1, answer: 'B', timeUsed: 2}, {_id: 2, answer: 'A', timeUsed: 2}, {_id: 3, answer: 'A', timeUsed: 5}], 
  result: { score: 3 } },
{ title: "Exam A",
  answer: [{_id: 1, answer: 'B', timeUsed: 4}, {_id: 2, answer: 'A', timeUsed: 5}, {_id: 3, answer: 'A', timeUsed: 9}], 
  result: { score: 1 } },
{ title: "Exam A", 
  answer: [{_id: 1, answer: 'B', timeUsed: 1}, {_id: 2, answer: 'A', timeUsed: 1}, {_id: 3, answer: 'A', timeUsed: 11}], 
  result: { score: 2 } }

You already answer how to config to find the avg score across records grouped by "createdAt".
How about configuration for finding avg "timeUsed" across records grouped by index "_id" in array.
My expected result would be something like this.

{ avgTimeUsedOfId1: 2.33, // (2 + 4 + 1) / 3
  avgTimeUsedOfId2: 2.66, // (2 + 5 + 1) / 3
  avgTimeUsedOfId3: 8.33 // (5 + 9 + 11) / 3}

Thank you in advance,
Steve