Merge pull request #10 from USEPA/master

Update from recipe

examples/traci.py

@@ -1,7 +1,5 @@
 import logging as log
 import os
-from os.path import join
-import yaml
 
 import lciafmt
 
@@ -10,11 +8,6 @@ def main():
         os.path.realpath(__file__)).replace('\\', '/')
     outputpath = modulepath + '/../output/'
     os.makedirs(outputpath, exist_ok=True)
-    datapath = modulepath + '/../lciafmt/data/'
-
-    with open(join(datapath, "TRACI_description.yaml")) as f:
-        metadata=yaml.safe_load(f)
-    method_description = metadata['description']
 
     log.basicConfig(level=log.INFO)
     data = lciafmt.get_method(lciafmt.Method.TRACI)
@@ -29,7 +22,7 @@ def main():
     json_pack = outputpath+"traci_2.1_json.zip"
     if os.path.exists(json_pack):
         os.remove(json_pack)
-    lciafmt.to_jsonld(mapped_data, json_pack, method_description)
+    lciafmt.to_jsonld(mapped_data, json_pack)
 
 
 

lciafmt/__init__.py

@@ -40,10 +40,10 @@ def clear_cache():
     cache.clear()
 
 
-def to_jsonld(df: pd.DataFrame, zip_file: str, description: str="", write_flows=False):
+def to_jsonld(df: pd.DataFrame, zip_file: str, write_flows=False):
     log.info("write JSON-LD package to %s", zip_file)
     with jsonld.Writer(zip_file) as w:
-        w.write(df, description, write_flows)
+        w.write(df, write_flows)
 
 
 def map_flows(df: pd.DataFrame, system=None, mapping=None,

lciafmt/data/ReCiPe2016_description.yaml

@@ -0,0 +1,57 @@
+description:
+    'ReCiPe2016 is built from LCIA Formatter v0.3-beta and mappings from the Federal 
+    Elementary Flow List (FEDEFL) v1.0.2
+    
+    ReCiPe 1.1 source file: https://www.rivm.nl/sites/default/files/2018-11/ReCiPe2016_CFs_v1.1_20180117.xlsx
+    
+    Source citation: (Huijbregts 2016)
+    
+    ReCiPe 1.1 flowable and context input files are maintained in the FEDEFL GitHub Repository: 
+    https://github.com/USEPA/Federal-LCA-Commons-Elementary-Flow-List 
+    
+   
+    The LCIA Formatter (https://github.com/USEPA/LCIAformatter) applies ReCiPe 1.1 
+    characterization factors to flows using the mapping housed in the FEDEFL. 
+    The flowable and context input files for ReCiPe 1.1 are used by the LCIA Formatter 
+    to translate the names from the ReCiPe 1.1 source file and generate a mapping 
+    file for the FEDEFL. This file contains all of the accepted flowable and context 
+    combinations which can be mapped to ReCiPe 1.1 characterization factors. It also 
+    includes target UUID, units, and conversion factors as well as tracking 
+    who mapped and verified each record and when they were last updated.
+
+    
+    FEDEFL contexts are mapped to ReCiPe 1.1 to enable impact assessment across most 
+    of the possible contexts available within FEDEFL. Mappings that occur in the 
+    primary contexts (e.g. air) are assigned to all possible sub-contexts which apply. 
+    Where a context is non-specific in FEDEFL relative to the available ReCiPe 1.1 
+    contexts (e.g. rural/urban), the LCIA Formatter applies the average of the 
+    relevant characterization factors from ReCiPe 1.1.
+    
+    
+    The LCIA Formatter generates a separate impact assement method for each of the 
+    three approaches employed in ReCiPe 1.1, which includes Individualist, 
+    Hierarchist, and Egalitarian. More details on the methods can be found on the 
+    website https://www.rivm.nl/en/life-cycle-assessment-lca/downloads
+
+
+    This mapping is representative of ReCiPe 1.1 midpoint impact factors and 
+    endpoint conversion factors are not included here.  
+    '
+
+ReCiPe 2016 - Midpoint/I:
+    '
+    
+    ReCiPe 2016 v1.1 midpoint method, Individualist version. The typical ReCiPe 
+    midpoint method used is the Hierarchist version.'
+
+ReCiPe 2016 - Midpoint/H:   
+    '
+    
+    ReCiPe 2016 v1.1 midpoint method, Hierarchist version. This is the typical 
+    ReCiPe midpoint method. ' 
+
+ReCiPe 2016 - Midpoint/E:
+    '
+    
+    ReCiPe 2016 v1.1 midpoint method, Egalitarian version. The typical ReCiPe 
+    midpoint method used is the Hierarchist version.'

lciafmt/data/ReCiPe2016_split.csv

@@ -0,0 +1,5 @@
+New Flowable,CAS
+"Chlordane, technical",12789036
+"FENOXYCARB, non-preferred CAS",79127803
+"FENPROPATHRIN, non-preferred CAS",64257847
+"Mecoprop, non-preferred CAS",93652

lciafmt/data/TRACI_description.yaml

@@ -5,6 +5,8 @@ description:
     TRACI 2.1 source file: 
     https://www.epa.gov/sites/production/files/2015-12/traci_2_1_2014_dec_10_0.xlsx
     
+    Source citation: (Bare 2012)
+    
     TRACI 2.1 flowable and context input files are maintained in the FEDEFL GitHub Repository: 
     https://github.com/USEPA/Federal-LCA-Commons-Elementary-Flow-List 
     

lciafmt/jsonld.py

@@ -6,7 +6,7 @@
 import olca.pack as pack
 import pandas
 
-from .util import make_uuid, is_non_empty_str
+from .util import make_uuid, is_non_empty_str, get_method_metadata
 
 
 class Writer(object):
@@ -25,9 +25,9 @@ def __enter__(self):
     def __exit__(self, exc_type, exc_val, exc_tb):
         self.__writer.close()
 
-    def write(self, df: pandas.DataFrame, description, write_flows=False):
+    def write(self, df: pandas.DataFrame, write_flows=False):
         for _, row in df.iterrows():
-            indicator = self.__indicator(row, description)
+            indicator = self.__indicator(row)
             factor = olca.ImpactFactor()
             flow = self.__flow(row)
             unit = row[8]
@@ -49,7 +49,7 @@ def write(self, df: pandas.DataFrame, description, write_flows=False):
             for v in d.values():
                 self.__writer.write(v)
 
-    def __indicator(self, row, description) -> olca.ImpactCategory:
+    def __indicator(self, row) -> olca.ImpactCategory:
         uid = row[3]
         if not is_non_empty_str(uid):
             uid = make_uuid(row[0], row[2])
@@ -65,19 +65,19 @@ def __indicator(self, row, description) -> olca.ImpactCategory:
         ind.impact_factors = []
         self.__indicators[uid] = ind
 
-        method = self.__method(row, description)
+        method = self.__method(row)
         ref = olca.ImpactCategoryRef()
         ref.id = uid
         ref.name = ind.name
         ref.ref_unit = ind.reference_unit_name
         method.impact_categories.append(ref)
         return ind
 
-    def __method(self, row, description) -> olca.ImpactMethod:
+    def __method(self, row) -> olca.ImpactMethod:
         uid = row[1]
         if not is_non_empty_str(uid):
             uid = make_uuid(row[0])
-
+        description = get_method_metadata(row[0])
         m = self.__methods.get(uid)
         if m is not None:
             return m

lciafmt/recipe.py

@@ -2,6 +2,7 @@
 
 import pandas
 import xlrd
+import os
 
 import lciafmt.cache as cache
 import lciafmt.df as df
@@ -25,6 +26,8 @@
         'sea water' : 'water/sea water',
         'Sea water' : 'water/sea water',
         'marine water' : 'water/sea water'}
+datapath = os.path.dirname(os.path.realpath(__file__)).replace('\\', '/')+'/data'
+flowables_split = pandas.read_csv(datapath+'/ReCiPe2016_split.csv')
 
 def get(add_factors_for_missing_contexts=True, file=None, url=None) -> pandas.DataFrame:
     log.info("get method ReCiPe 2016")
@@ -39,6 +42,21 @@ def get(add_factors_for_missing_contexts=True, file=None, url=None) -> pandas.Da
     if add_factors_for_missing_contexts:
         log.info("Adding average factors for primary contexts")
         df = util.aggregate_factors_for_primary_contexts(df)    
+
+    log.info("Handling manual replacements")
+    """ due to substances listed more than once with the same name but different CAS
+    this replaces all instances of the Original Flowable with a New Flowable
+    based on a csv input file according to the CAS"""
+    for index, row in flowables_split.iterrows():
+        newCAS = util.format_cas(row['CAS'])
+        newFlow = row['New Flowable']
+        df.loc[df['CAS No'] == newCAS, 'Flowable'] = newFlow
+
+    length=len(df)
+    df.drop_duplicates(keep='first',inplace=True)
+    length=length-len(df)
+    log.info("%s duplicate entries removed", length)
+
     return df
 
 
@@ -86,7 +104,7 @@ def _read_mid_points(sheet: xlrd.book.sheet, records: list):
                 flow_unit = flow_unit.split("/")[1].strip()
         cas = ""
         if cas_col > -1:
-            util.format_cas(xls.cell_str(sheet, row, cas_col))
+            cas=util.format_cas(xls.cell_f64(sheet, row, cas_col))
 
         if with_perspectives:
             for i in range(0, 3):
@@ -134,6 +152,9 @@ def _find_data_start(sheet: xlrd.book.sheet) -> (int, int, bool):
 
 
 def _find_flow_column(sheet: xlrd.book.sheet) -> int:
+    if _containstr(sheet.name, "land", "occupation"):
+        ncol = 1
+        return ncol
     ncol = -1
     for row, col in xls.iter_cells(sheet):
         s = xls.cell_str(sheet, row, col)
@@ -227,7 +248,7 @@ def _determine_compartments(sheet: xlrd.book.sheet) -> (str, int):
             break
         s = xls.cell_str(sheet, row, col)
         if _containstr(s, "compartment") \
-            or _containstr(s, "name", "in", "ecoinvent"):
+            or _containstr(s, "name", "in", "ReCiPe"):
             compartment_col = col
             break
 

lciafmt/util.py

@@ -1,6 +1,10 @@
 import uuid
 import pandas as pd
 import numpy as np
+import yaml
+import os
+from os.path import join
+import logging as log
 
 def make_uuid(*args: str) -> str:
     path = _as_path(*args)
@@ -97,3 +101,23 @@ def aggregate_factors_for_primary_contexts(df) -> pd.DataFrame:
     df = pd.concat([df, df_secondary_agg], ignore_index=True, sort=False)
     return df
 
+def get_method_metadata(name: str) -> str:
+    modulepath = os.path.dirname(
+    os.path.realpath(__file__)).replace('\\', '/')
+    datapath = modulepath + '/../lciafmt/data/'
+    if name == "TRACI 2.1": 
+        method = 'TRACI'
+    elif "ReCiPe 2016" in name:
+        method = 'ReCiPe2016'
+    else:
+        return ""
+    with open(join(datapath, method + "_description.yaml")) as f:
+        metadata=yaml.safe_load(f)
+    method_description = metadata['description']
+    detail = ""
+    try:
+        detail = metadata[name]
+        method_description = method_description+detail
+    except:
+        log.info("No further detail in description")
+    return method_description

setup.py

@@ -2,7 +2,7 @@
 
 setup(
     name="lciafmt",
-    version="0.2-beta",
+    version="0.3-beta",
     packages=["lciafmt"],
     install_requires=["fedelemflowlist @ git+git://github.com/USEPA/Federal-LCA-Commons-Elementary-Flow-List@v1.0.1.1#egg=fedelemflowlist",
                       "olca-ipc", "pandas", "xlrd"],