pythonpostgresqlolapolap-cubecubes

Python Cubes Olap Framework Date PointCut


So i am trying to implement some BI with Python Cubes Framework and i am running into some problems.

Basically i am trying to do a "simple" PointCut/slice and dice and i am not having any luck with it. I am using a PostgreSQL database with PostGis.

My model.json is:

{
    "dimensions": [
        {"name": "user", "attributes": ["id", "username"]},
        {"name": "resources", "attributes": ["id", "resource_simple_name"]},
        {"name":"created_on", "role": "time"}
    ],
    "cubes": [
        {
            "name": "users_resources_likes",
            "dimensions": ["user", "resources", "created_on"],
            "mappings": {
                "user.id": "auth_user.id",
                "user.username": "auth_user.username",
                "resources.id": "resources.id",
                "resources.resource_simple_name": "resources.resource_simple_name",
                "created_on": "created_on"
            },
            "joins": [
                {
                    "master": "user_id",
                    "detail": "auth_user.id"
                },
                {
                    "master": "resource_id",
                    "detail": "resources.id"
                }
            ]
        }

    ]
}

if i try do a poincut with date on slicer

aggregate?drilldown=created_on&cut=created_on:2012

i get a DataError: (DataError) invalid input syntax for type timestamp with time zone: "2012"

After searching a while i read that it could be because my postgresql database has a timestamp:

created_on timestamp with time zone NOT NULL DEFAULT '2014-02-10 00:00:00+00'::timestamp with time zone,

SO i tried to do:

?drilldown=created_on&cut=created_on:2012-09-15T09:37:59+00:00

and i get a :

{
error: "unknown_user_error",
message: "Wrong dimension cut string: 'created_on:2012-09-15T09:37:59 00:00'"
}

What am i doing wrong? is the problem in my model.json?


Solution

  • I found that adding more information to the date in the model solved the problem:

        {
            "name": "created_on",
            "label": "Date Created",
            "role": "time",
            "info": {
                "cv-datefilter": true,
                "cv-datefilter-hierarchy": "weekly"
            },
            "levels": [
                   {
                       "name":"year",
                       "label":"Year",
                       "info": { "cv-datefilter-field": "year" }
                   },
                   {
                       "name":"quarter",
                       "label":"Quarter"
                   },
                   {
                       "name":"month",
                       "label":"Month"
                   },
                   {
                       "name":"week",
                       "label":"Week",
                       "info": { "cv-datefilter-field": "week" }
                   }
               ],
            "hierarchies": [
                {
                    "name": "weekly",
                    "label": "Weekly",
                    "levels": [ "year", "week"]
                },
                {
                    "name": "monthly",
                    "label": "Monthly",
                    "levels": [ "year", "quarter", "month"]
    
                }
            ]
        }
    

    and the mappings:

            "mappings": {
                "created_on.year": {"column":"created_on", "extract":"year"},
                "created_on.month": {"column":"created_on", "extract":"month"},
                "created_on.week": {"column":"created_on", "extract":"week"},
                "created_on.day": {"column":"created_on", "extract":"day"},
                "created_on.quarter": {"column":"created_on", "extract":"quarter"}
            }