Lists the IDs, name, and descriptions of all metrics.
Example output:
{
"metrics": [
{
"id": 1,
"name": "CPU Architecture",
"description": "The architecture of the client machine"
"tooltip": "{y}% of all {pop} records have {x} architecture",
"type": "categorical"
},
{
"id": 2,
"name": "CPU Count",
"description": "The number of CPUs available"
"tooltip": "{p}% of all {pop} records have {x} or more CPUs",
"type": "numerical"
},
...
]
}
Returns data for a specific metric, grouped by the population, if provided.
Query parameters:
date
(optional): A date in ISO 8601 format. This date will be used to find the most recent prior data set.pop
(optional): A comma separated list of populations. If none provided, this will default to "All".
Example output:
{
"id": 1,
"dataSet": "2016-01-01",
"type": "categorical",
"name": "CPU Architecture",
"populations": [
{
"name": "All",
"numObs": 5321560,
"points": [
{
"p": 0.932504,
"refRank": 1,
"b": "x86",
"c": 0.932504
},
{
"p": 0.0672972,
"refRank": 2,
"b": "x86-64",
"c": 0.999801
},
{
"p": 0.00015503,
"refRank": 3,
"b": "missing",
"c": 0.999956
},
{
"p": 0.0000362676,
"refRank": 4,
"b": "arm",
"c": 0.999992
},
{
"p": 0.00000357038,
"refRank": 5,
"b": "sparc",
"c": 0.999996
},
{
"p": 0.00000432204,
"refRank": 6,
"b": "ppc",
"c": 1
}
]
}
]
}
- dataSet
-
All metrics are connected to a
DataSet
which is the date the data was created and imported. This allows for historicalDataSet
s to exist. - populations.name
- The population this data set applies to. Besides "All", some examples include: "channel:release" or "os:darwin".
- type
-
One of
categorical
ornumerical
. When the type iscategorical
,points.b
will be a string containing a word like arm or sparc. When the type isnumerical
,points.b
will be a string containing a floating-point number. - points
- b
- The bin to which a given amount of weight is assigned
<dt>p</dt>
<dd>
The proportion of weight assigned to that bin (all <code>p</code>s should
sum to 1)
</dd>
<dt>c</dt>
<dd>
The cumulative amount of weight if the bins are sorted in
<code>refRank</code> order
</dd>
<dt>refRank</dt>
<dd>
The sort order of the bin from the reference population ordered from most
weight to least weight. This is needed because if you display more than one
population, the proportion of records bins will be be different, so you need
a stable ranking to display them side-by-side.
</dd>
- Data points are guaranteed to be in order, such that the
c
and theb
(when the data is numerical) of an object are always higher than those of the object before it.