This Python module consists of a collection of utilities for importing and translating hazard data from publicly available Internet resources, translating to an intermediate GreenScreen datastructure and computing an overall GreenScreen benchmark score according to the GreenScreen for Safer Chemicals methodology.
The entire benchmarking process, including the initial import of data, translation to GreenScreen representation, and assessment of an overall benchmark score, is completely computer automated. The resulting data is exported in a JavaScript Object Notation (JSON) file format and is suitable to be uploaded to a document storage system or relational database for later retrieval.
At present, the module supports loading data from two hazard endpoint data sources:
-
Chemical Management Center, part of a Japanese governmental organization called the National Institute for Technology Evaluation. In the source code, the dataset is referred to as the GHS Japan country list and is available at the following url: http://www.safe.nite.go.jp/english/ghs/ghs_index.html. According to the GreenScreen methodology, this list is classified as a Screening A list.
-
Proposition 65, a list of chemicals known by the State of California to have a 1 in 100,000 chance of causing cancer, birth defects or developmental harm. The dataset is available at the following url: http://www.oehha.ca.gov/prop65/prop65_list/Newlist.html According to the GreenScreen methodology, this list is classified as an Authoritative A list.
The benchmark scores computed by this module do not constitute comprehensive GreenScreen assessments. In addition to a benchmark score, a comprehensive GreenScreen assessment contains a detailed report and verification by a GreenScreen licensed profiler. Furthermore, a comprehensive GreenScreen assessment is carried out on the basis of hazard endpoints spanning most, if not all, of the eighteen categories described in the GreenScreen for Safer chemicals methodology.
In this module's current state of development, a benchmark score is computed only on the basis of data imported from the hazard endpoint data sources listed above. Therefore, any benchmark scores computed by the module must be considered provisional in nature, as appending additional hazard endpoint data may result in a lower (more hazardous) overall benchmark score.
Nonetheless, the module is effective to identify Benchmark 1, Avoid: Chemical of High Concern chemicals, which, according the GreenScreen ListTranslator v1.2 method, are reported as LT-1: Benchmark 1 or LT-P1: Possible Benchmark 1 scores, depending on the source of the hazard endpoint that resulted in a provisional Benchmark 1 assessment. All other provisional scores are reported as LT-U: Unspecified Benchmark.
Before using the module, it is recommended to register with the nonprofit group, Clean Production Action, to receive specific details on performing GreenScreen assessments and to familiarize yourself with the benchmarking and translation process. http://www.greenscreenchemicals.org/method/?/Greenscreen.php
As discussed above, as the results of the automated benchmarking process do not constitute official, comprehensive assessments, the overall assessment should be reported to a general audience as one of the following:
- LT-1: Benchmark 1
- LT-P1: Possible Benchmark 1
- LT-U: Unspecified Benchmark
During the benchmarking process, a provisional benchmark score is computed and stored in the data structure. The possible provisional benchmark scores are listed below:
- Provisional Benchmark 1, Avoid: Chemical of High Concern
- Provisional Benchmark 2, Use: But Search for Safer Substitutes
- Provisional Benchmark 3, Use: But Still Opportunity for Improvement
- Provisional Benchmark 4, Prefer - Safer Chemical
- Provisional Benchmark U, Unknown
The provisional scores should not be reported to a general audience, but are recorded during the benchmarking process for their value to GreenScreen practitioners. For example, once a licensed profiler deems that sufficient hazard endpoint data has been appended for a given chemical and a corresponding report has been generated, a provisional score could later be reported as a comprehensive score.
After cloning the project, install the Python module to your system path:
python setup.py installA batch processing script is provided that will download data from the GHS-Japan website, perform list translation and benchmarking. After installing the Python module in the previous step, the greenscreen_batch_process command line utility will be available and can be executed by calling:
greenscreen_batch_process <data directory>Replace <data directory> with the directory you wish to store the
information.
Alternatively, any of the functions contained within the module can be reused in other Python applications.
The greenscreen_framework module class definitions can then be used in other Python programs by putting the following at the top of your Python script:
import greenscreen_framework.greenscreen as gs
import greenscreen_framework.ghs as ghs
import greenscreen_framework.prop65 as prop65Refer to the batch processing script greenscreen_batch_process for example usage.
All notable changes to this project will be documented below.
- prop65.py contains a Prop65 python class to load Proposition 65 data from http://www.oehha.ca.gov/prop65/prop65_list/Newlist.html and translate/save the results into an intermediate JSON file.
- greenscreen.py added a method to load/merge Prop65 data that has been translated/saved in an intermediate JSON format.
- README.md updated description of project
- Initial commit
Copyright 2013-2016 Kristopher Wehage