The following database rules explain the usage of the database called oedb, which will become a part of the openmod internet presence. For further details see also here.
Data Documentation
- All data included in the databases has be documented! On this wiki page you learn, how to do so.
- All abreviations have to be documentated in the !
Naming of Data
The data in the database is organised in schemata and tables, which names are important to find around.
Database Name
The name of the database is oedb.
Database Schema
The structure of the database is realised via the naming of the schemata, which follows the following rules:
- always lower case
- no points, no commas
- no spaces
- no dates
- use underscores
- The name starts with the type of the schema:
- orig for original data: Data that is directly available in other sources and underwent no editing other than transfering it to a data base friendly format
- calc for processed data: Data that origin from original data or processed data via some script or other modifications. The specific data and methods used for this computations must be mentioned.
- res for result data: Data that is the result of an energy model
- The name includes the distinct subject area or source.
Example: orig_vg250
Database Table
- always lower case
- no points, no commas
- no spaces
- no dates
- use underscores
The table name consists of
- name starts with the source (e.g. zensus)
- main value (e.g. population)
- if separated by [attribute] (e.g. by_gender)
- with resolution [tupel] (e.g. per_mun)
Example: zensus_population_by_gender_per_mun
Data Integrity
Data Integrity is one aspect of ensuring data quality the in the oedb.
General
- Each table should have a single auto-incremented integer column as Primary Key with the corresponding table constraint.
- Grant all privilegs to oeuser
Geografic Data
- WGS84 - EPSG: 4326 (http://spatialreference.org/ref/epsg/4326/)
- ETRS89 / ETRS-LAEA - EPSG: 3035 (http://spatialreference.org/ref/epsg/3035/)
Data Referencing
There are several usefull attributes on a table users have to know in order to work with the contained data efficiently. These informations are separated in two major categories:
- Comments on tables: Contains metadata that define several aspects for the whole table
- Comments on rows: Store important informations on a specific row and the stored data
Comments on a table are json-dictionary as strings that describe the overall structure of the corresponding table. For every field the corresponding unit (in case of numerical values) and a description that specifies the contained data should be attached.
Original Data (orig)
{"Name": "The Full Name",
"Source": ["Name", "www.website.com / registation required"],
"Reference date": ["2013"],
"Date of collection": ["01.08.2013"],
"Original file": ["346-22-5.xls"],
"Spatial resolution": ["Germany"],
"Description": ["Example Data (annual totals)", "Regional level: national"],
"Table fields": [
{"name":"id",
"description"Unique identifier"",
"description_german":"",
"unit":"" },
{"name":"year",
"description"Reference Year"",
"description_german":"",
"unit":"" },
{"name":"example_value",
"description"Some important value"",
"description_german":"",
"unit":"EUR" }],
"Changes":[
{"name":"Joe Nobody",
"mail":"joe.nobody@gmail.com (fake)",
"date":"16.06.2014",
"comment":"Created table" },
{"name":"Joana Anybody",
"mail":"joana.anybody@gmail.com (fake)",
"date":"17.07.2014",
"comment":"Translated field names"}],
"ToDo": ["Some datasets are odd -> Check numbers against another data"],
"Licence": ["Licence – Version 2.0 (dl-de/by-2-0; [http://www.govdata.de/dl-de/by-2-0])"],
"Instructions for proper use": ["Always state licence"]}
Processed Data (calc)
As processed data does not necessarily have a simple source the fields "Source", "Original File", "Reference date" are ommited. These information should be mentioned row-wise as explained in the next paragraph.
{"Name": "Results",
"Date of collection": ["01.08.2013"],
"Spatial resolution": ["Germany"],
"Description": ["Financial key figures of German municipalities (annual totals)",
"Regional level: municipalities, association of municipalities"],
"Table fields": [
{"name":"id",
"description":"Unique identifier,
"description_german":"",
"unit":"" },
{"name":"year",
"description":"Reference Year",
"description_german":"",
"unit":"" },
{"name":"example_value",
"description":"Some important value",
"description_german":"",
"unit":"EUR" }],
"Changes":[
{ "name":"Autor1",
"mail":"Autor1@e-mail.com",
"date":"16.06.2014",
"comment":"Created table" },
{ "name":"Autor2",
"mail":"Autor2@e-mail.com",
"date":"17.07.2014",
"comment":"Translated field names"}],
"ToDo": ["Some datasets are odd -> Check numbers against another data"],
"Licence": ["Datenlizenz Deutschland – Namensnennung – Version 2.0 (dl-de/by-2-0; [http://www.govdata.de/dl-de/by-2-0 http://www.govdata.de/dl-de/by-2-0])"],
"Instructions for proper use": ["Always state licence"]}
The comments on a row aim to enhance the reproducabilty of data that is present in the database by stating the underlying data that was used to compute this row as well as the applied computational methods.
During these processes some kinds of uncertainities might occur as well as the decissions (assumtions) that were made to deal with them. Examplatory, some values might be missing amid this row and a linear development during this period was assumed to fix this gap. Each row has to be annotated by a json dictionary that must contain the following fields:
- origin: Link or textual description of the data set this row origins from.
- method: Method used to calculate this row from above origin (e.g. Link to a python script)
- assumption: A list of dictionaries. Each dictionary describes an assumption and annotates the affected rows.
- begin: First column affected by the assumption
- end: Last column affected by the assumption
- type: Type of the problem that had to be solved. Each type requires one or more additional keys in this dictionary. Possible types and their required additional keys are:
- gap: A not all fields could be calculated and/or filled
- solution: Method that was used to generate date to fill this gap (e.g. linear interpolation)
- multiplicity: A field could be filled by several values
- solution: Method that was used to select one value (e.g. Minimum)
- values: Possible Values that could have been used
An examplatory dictionary:
{
"origin":"https://data.openmod-initiative.org/data/oedb/orig_db/table
"method":"https://github.com/openego/data_processing/blob/master/calc_ego_substation/Voronoi_ehv.sql"
"assumptions":
[
{
"type": "gap"
"begin": "step_15"
"end": "step_34"
"solution": "linear_interpolation"
}
]
}
