Actually I'm using Python with RPY2 to manipulate data and GG Plot to create beautiful graphics. . I have some data in the PostgreSQL database, and I am using psycopg2 to query the data.
I am starting a thesis, and in future I need to store my (very large) simulation data: multiple dimensions, aggregation queries, etc.
Is there a best or standard practice for the interface between Python (and I do not want Python + R, java, or any other dashboard in Java) and like the OLAP engine? I searched Google for any solution, and I could not find anything.
I have assessed in a nutshell, and Django-ORM, but they have OLAP server (Madrians or other) ...
Is it possible to write a question in MDX, and psycopg With ODBC, query my obsolete server, and the olap server is answering me with my simulation data (no mapping is the python object, but is it ok for me?)
Update 1:
I need to search around the OLAP + Mondrian Thus technology?
Because (GeoSoa Department + Thierry Birdard) wrote a spatial extension for OLAP: Solap, and implemented it in the Mondrian. This interest is because I'm working on spatial multi-agent based simulation (~ = geosimulation).
The GeoSoa Department has created an Ajax-based component that visualizes and visualizes spatial data with geographic communication: Soulplayer, which can do a query by the Mondrian server through its Xlma servlet.
Problem: Probably large data manipulation requires slow speed, Internet or Apache. In essence, this is only to visualize data or map ... In my case, I need raw data to manipulate your own data + graphics with R: spatial analysis, regression analysis, rank tail, etc. Here, SOLAP helps me in preparing data for complex R analysis later on.
Why Python?
1 - Web access to spatial data -
I am trying to use the "cool" Python framework, such as: big gis , OpenSource includes visualization with community, local manipulation of queries / data manipulation, JavaScript extensions and more.
2 - Local access to spatial data in GIS -
I want to create a plugin in QGIS (Open Source GIS), and QGIS plugin for data access and visualization API = Python.
3 - Automatic analysis of data -
Runs a simulation with a user or a scientific grid Select computing and automatic analysis (R + ggplot2 + MDX query) Want to run on data? My goal is to create a synthetic report of simulation (graphic, tabular data, etc.).
Therefore, after the simulation, the data goes to the OLAP / SOP cube, and many Python script users), get data with MDX, manipulate data with R + RPY2, and doc-wiki or Write and write good output to the scientists on any other community-platform.
The problem?
2 - Sawaflayers uses Ajax to access data, too slow for me is.
3 - SQLLame and GeoLee do not have any driver connection for a multi-dimensional database (OLAP).
* Solution? *
1 - Py4j with Python to access Java object or Java collection in olap4j? Write your own work to access the Java mapped archive? => Dangerous and not very easy? ...
2 - XLMA with Ajax Mondrian Server? It is very slow.
3 - Write your own pin-connector for OLP Mondrian? => Ouch it's a difficult way, I guess.
What should I do?
As you know, Mondrian is a complete OLP engine that is the top of the database written in Java But if I understand your question, then you want to use the Mondrian and wonder how it can be interfaced with Python.
I have to process the MDX queries on the command line and. Python calls it directly to the command line
import order results = commands.getoutput ('java-jor Mondrian_cli.jar -q select non Imperial crossjeno ({[remedy]. [ Store sales]}, crosgen ([time]. [1997]. Children, [store]. [All stores] children)), [Product]. [All products]. From sales [sales] to hair And for server usage, I package it in servlet and I send MDX with AJAX. Ajax calls are not a big overhead and that is why I do not need Python and Java coupling rather than communicating with the Mondrian server.
Comments
Post a Comment