Added processing to apply project credits, determine institution name…

… for each PI, and exporting HU and BU invoices

process_report/institute_map.json

@@ -0,0 +1,27 @@
+{
+    "northeastern.edu"      : "Northeastern University",
+    "bu.edu"                : "Boston University",
+    "bentley.edu"           : "Bentley",
+    "uri.edu"               : "University of Rhode Island",
+    "redhat.com"            : "Red Hat",
+    "childrens.harvard.edu" : "Boston Childrens Hospital",
+    "mclean.harvard.edu"    : "McLean Hospital",
+    "meei.harvard.edu"      : "Massachusetts Eye & Ear",
+    "dfci.harvard.edu"      : "Dana-Farber Cancer Institute",
+    "bwh.harvard.edu"       : "Brigham and Women's Hospital",
+    "bidmc.harvard.edu"     : "Beth Israel Deaconess Medical Center",
+    "harvard.edu"           : "Harvard University",
+    "wpi.edu"               : "Worcester Polytechnic Institute",
+    "mit.edu"               : "Massachusetts Institute of Technology",
+    "umass.edu"             : "University of Massachusetts Amherst",
+    "uml.edu"               : "University of Massachusetts Lowell",
+    "codeforboston.org"     : "Code For Boston",
+    "mmsh"                  : "Harvard University",
+    "gstuart"               : "University of Massachusetts Amherst",
+    "rudolph"               : "Boston Childrens Hospital",
+    "robbaron"              : "Boston University",
+    "kmdalton"              : "Harvard University",
+    "mzink"                 : "University of Massachusetts Amherst",
+    "yale.edu"              : "Yale University",
+    "francesco.pontiggia"   : "Harvard University"
+}

process_report/process_report.py

@@ -1,9 +1,101 @@
 import argparse
 import os
+import sys
 
+import json
 import pandas
 
 
+### Invoice field names
+INVOICE_DATE_FIELD = 'Invoice Month'
+PROJECT_FIELD = 'Project - Allocation'
+PROJECT_ID_FIELD = 'Project - Allocation ID'
+PI_FIELD = 'Manager (PI)'
+INVOICE_EMAIL_FIELD = 'Invoice Email'
+INVOICE_ADDRESS_FIELD = 'Invoice Address'
+INSTITUTION_FIELD = 'Institution'
+INSTITUTION_ID_FIELD = 'Institution - Specific Code'
+SU_HOURS_FIELD = 'SU Hours (GBhr or SUhr)'
+SU_TYPE_FIELD = 'SU Type'
+COST_FIELD = 'Cost'
+CREDIT_FIELD = 'Credit'
+CREDIT_CODE_FIELD = 'Credit Code'
+BALANCE_FIELD = 'Balance'
+###
+
+
+def apply_credits_0001(dataframe):
+    credit_code = "0001"
+    pass
+
+
+def apply_credits_0002(dataframe):
+    """Applies the New PI Credit. This credit function expects the 
+    env var `C0002_OLD_PI` to be set, pointing to a txt file containing old PIs"""
+    credit_code = "0002"
+    credit_amount = 1000
+
+    old_pi_list = set()
+    if "C0002_OLD_PI" not in os.environ:
+        print("Applying credit 0002 failed. C0002_OLD_PI env var is not set")
+        sys.exit(1)
+
+    try:
+        with open(os.getenv("C0002_OLD_PI")) as f:
+            for pi in f: old_pi_list.add(pi.strip())
+
+    except FileNotFoundError:
+        print("Applying credit 0002 failed. Old PI file does not exist")
+        sys.exit(1)
+
+    print("Old pi list: ", old_pi_list)
+    pi_list = dataframe[PI_FIELD].unique()
+
+    for pi in pi_list:
+        if pandas.isna(pi): 
+            continue   # NaN check
+        if pi in old_pi_list: 
+            continue  # Is the PI an old PI?
+
+        pi_projects = dataframe[dataframe[PI_FIELD] == pi]
+        rem_credit = credit_amount
+        for i, row in pi_projects.iterrows():
+            project_cost = row[COST_FIELD]
+            if project_cost >= rem_credit:
+                dataframe.at[i, CREDIT_FIELD] =  rem_credit
+                dataframe.at[i, CREDIT_CODE_FIELD] = credit_code
+                dataframe.at[i, BALANCE_FIELD] = row[COST_FIELD] - rem_credit
+                break
+            else:
+                dataframe.at[i, CREDIT_FIELD] =  project_cost
+                dataframe.at[i, CREDIT_CODE_FIELD] = credit_code
+                dataframe.at[i, BALANCE_FIELD] = 0
+                rem_credit -= project_cost
+
+    return dataframe
+
+
+applied_credits = [apply_credits_0002]
+
+
+def get_institution_from_pi(pi_uname):
+
+    dir_path = os.path.dirname(__file__)
+    with open(f'{dir_path}/institute_map.json', 'r') as f:
+        institute_map = json.load(f)
+
+    if '@' in pi_uname:
+        domain = pi_uname.split('@')[1]
+        institute_name = institute_map.get(domain, '')
+    else:
+        institute_name = institute_map.get(pi_uname, '')
+
+    if institute_name == '':
+            print(f"PI name {pi_uname} does not match any institution!")
+
+    return institute_name
+
+
 def main():
     """Remove non-billable PIs and projects"""
 
@@ -41,6 +133,18 @@ def main():
         default="pi_invoices",
         help="Name of output folder containing pi-specific invoice csvs"
     )
+    parser.add_argument(
+        "--HU-only",
+        required=False,
+        default="HU_only.csv",
+        help="Name of output csv for HU invoices"
+    )
+    parser.add_argument(
+        "--HU-BU",
+        required=False,
+        default="HU_BU.csv",
+        help="Name of output csv for HU and BU invoices"
+    )
     args = parser.parse_args()
     merged_dataframe = merge_csv(args.csv_files)
 
@@ -60,9 +164,13 @@ def main():
 
     projects = list(set(projects + timed_projects_list))
 
+    merged_dataframe = add_credits(merged_dataframe)
+    merged_dataframe = add_institution(merged_dataframe)
     billable_projects = remove_non_billables(merged_dataframe, pi, projects, args.output_file)
     remove_billables(merged_dataframe, pi, projects, "non_billable.csv")
     export_pi_billables(billable_projects, args.output_folder)
+    export_HU_only(billable_projects, args.HU_only)
+    export_HU_BU(billable_projects, args.HU_BU)
 
 
 def merge_csv(files):
@@ -83,7 +191,7 @@ def get_invoice_date(dataframe):
     Note that it only checks the first entry because it should
     be the same for every row.
     """
-    invoice_date_str = dataframe['Invoice Month'][0]
+    invoice_date_str = dataframe[INVOICE_DATE_FIELD][0]
     invoice_date = pandas.to_datetime(invoice_date_str, format='%Y-%m')
     return invoice_date
 
@@ -102,7 +210,7 @@ def timed_projects(timed_projects_file, invoice_date):
 
 def remove_non_billables(dataframe, pi, projects, output_file):
     """Removes projects and PIs that should not be billed from the dataframe"""
-    filtered_dataframe = dataframe[~dataframe['Manager (PI)'].isin(pi) & ~dataframe['Project - Allocation'].isin(projects)]
+    filtered_dataframe = dataframe[~dataframe[PI_FIELD].isin(pi) & ~dataframe[PROJECT_FIELD].isin(projects)]
     filtered_dataframe.to_csv(output_file, index=False)
     return filtered_dataframe
 
@@ -112,21 +220,60 @@ def remove_billables(dataframe, pi, projects, output_file):
 
     So this *keeps* the projects/pis that should not be billed.
     """
-    filtered_dataframe = dataframe[dataframe['Manager (PI)'].isin(pi) | dataframe['Project - Allocation'].isin(projects)]
+    filtered_dataframe = dataframe[dataframe[PI_FIELD].isin(pi) | dataframe[PROJECT_FIELD].isin(projects)]
     filtered_dataframe.to_csv(output_file, index=False)
 
+
 def export_pi_billables(dataframe: pandas.DataFrame, output_folder):
     if not os.path.exists(output_folder):
         os.mkdir(output_folder)
 
-    invoice_month = dataframe['Invoice Month'].iat[0]
-    pi_list = dataframe['Manager (PI)'].unique()
+    invoice_month = dataframe[INVOICE_DATE_FIELD].iat[0]
+    pi_list = dataframe[PI_FIELD].unique()
 
     for pi in pi_list:
-        pi_projects = dataframe[dataframe['Manager (PI)'] == pi]
-        pi_instituition = pi_projects['Institution'].iat[0]
+        if pandas.isna(pi): 
+            continue
+        pi_projects = dataframe[dataframe[PI_FIELD] == pi]
+        pi_instituition = pi_projects[INSTITUTION_FIELD].iat[0]
         pi_projects.to_csv(output_folder + f"/{pi_instituition}_{pi}_{invoice_month}.csv")
-
+
+
+def add_credits(dataframe : pandas.DataFrame):
+    """Adds credits to PIs depending on different criterions"""
+    dataframe.insert(dataframe.columns.get_loc(COST_FIELD) + 1, CREDIT_FIELD, 0.0)
+    dataframe.insert(dataframe.columns.get_loc(CREDIT_FIELD) + 1, CREDIT_CODE_FIELD, None)
+    dataframe.insert(dataframe.columns.get_loc(CREDIT_CODE_FIELD) + 1, BALANCE_FIELD, 0.0)
+
+    # Apply credits
+    for credit_func in applied_credits:
+        dataframe = credit_func(dataframe)
+
+    return dataframe
+
+
+def add_institution(dataframe: pandas.DataFrame):
+    """Determine the PI's institution name, logging any PI whose institution cannot be determined"""
+    for i, row in dataframe.iterrows():
+        pi_name = row[PI_FIELD]
+        if pandas.isna(pi_name): 
+            print(f"Project {row[PROJECT_FIELD]} has no PI") # Nan check
+        else: 
+            dataframe.at[i, INSTITUTION_FIELD] = get_institution_from_pi(pi_name)
+
+    return dataframe
+
+
+def export_HU_only(dataframe, output_file):
+    HU_projects = dataframe[dataframe[INSTITUTION_FIELD] == 'Harvard University']
+    HU_projects.to_csv(output_file)
+
+
+def export_HU_BU(dataframe, output_file):
+    HU_BU_projects = dataframe[(dataframe[INSTITUTION_FIELD] == 'Harvard University') | 
+                               (dataframe[INSTITUTION_FIELD] == 'Boston University')]
+    HU_BU_projects.to_csv(output_file)        
+
 
 if __name__ == "__main__":
     main()

process_report/tests/unit_tests.py

@@ -1,4 +1,5 @@
 from unittest import TestCase
+from unittest import skipIf
 import tempfile
 import pandas
 import os
@@ -175,3 +176,65 @@ def test_export_pi(self):
         self.assertNotIn('ProjectA', pi_df['Project - Allocation'].tolist())
         self.assertNotIn('ProjectB', pi_df['Project - Allocation'].tolist())
         self.assertNotIn('ProjectC', pi_df['Project - Allocation'].tolist())
+
+
+class TestGetInstitute(TestCase):
+
+    def setUp(self):
+
+        data = {
+            'Manager (PI)': ['[email protected]', '[email protected]', '[email protected]', 'fake', '[email protected]'],
+            'Project - Allocation': ['ProjectA', 'ProjectB', 'ProjectC', 'ProjectD', 'ProjectE'],
+            'Answer': ["Boston University", "McLean Hospital", "Harvard University", "", "Northeastern University"]
+        }
+        self.data = pandas.DataFrame(data)
+
+    def test_get_pi_institution(self):
+        for i, row in self.data.iterrows():
+            institution_name = process_report.get_institution_from_pi(row['Manager (PI)'])    
+            self.assertEqual(institution_name, row['Answer'])
+
+@skipIf(process_report.apply_credits_0002 not in process_report.applied_credits, "Skipping test for credit 0002 because credit not enabled")
+class TestCredit0002(TestCase):
+    def setUp(self):
+
+        data = {
+            'Manager (PI)': ['PI1', 'PI1', 'PI2', 'PI3', 'PI4', 'PI4'],
+            'Project - Allocation': ['ProjectA', 'ProjectB', 'ProjectC', 'ProjectD', 'ProjectE', 'ProjectF'],
+            'Cost': [10, 100, 10000, 5000, 800, 1000]
+        }
+        self.dataframe = pandas.DataFrame(data)
+        old_pi = ['PI2', 'PI3']
+        old_pi_file = tempfile.NamedTemporaryFile(delete=False, mode='w', suffix='.csv')
+        for pi in old_pi: old_pi_file.write(pi + "\n")
+        self.old_pi_file = old_pi_file.name
+
+        os.environ["C0002_OLD_PI"] = self.old_pi_file
+
+    def tearDown(self):
+        os.remove(self.old_pi_file)
+
+    def test_apply_credit_0002(self):
+        dataframe = process_report.add_credits(self.dataframe)
+
+        self.assertTrue('Credit' in dataframe)
+        self.assertTrue('Credit Code' in dataframe)
+        self.assertTrue('Balance' in dataframe)
+
+        credited_projects = dataframe[dataframe['Credit Code'] == '0002']
+
+        self.assertEqual(4, len(credited_projects.index))
+        self.assertTrue('PI2' not in credited_projects['Manager (PI)'].unique())
+        self.assertTrue('PI3' not in credited_projects['Manager (PI)'].unique())
+
+        self.assertEqual(10, credited_projects[credited_projects['Project - Allocation'] == 'ProjectA']['Credit'].iloc[0])
+        self.assertEqual(100, credited_projects[credited_projects['Project - Allocation'] == 'ProjectB']['Credit'].iloc[0])
+        self.assertEqual(800, credited_projects[credited_projects['Project - Allocation'] == 'ProjectE']['Credit'].iloc[0])
+        self.assertEqual(200, credited_projects[credited_projects['Project - Allocation'] == 'ProjectF']['Credit'].iloc[0])
+
+        self.assertEqual(0, credited_projects[credited_projects['Project - Allocation'] == 'ProjectA']['Balance'].iloc[0])
+        self.assertEqual(0, credited_projects[credited_projects['Project - Allocation'] == 'ProjectB']['Balance'].iloc[0])
+        self.assertEqual(0, credited_projects[credited_projects['Project - Allocation'] == 'ProjectE']['Balance'].iloc[0])
+        self.assertEqual(800, credited_projects[credited_projects['Project - Allocation'] == 'ProjectF']['Balance'].iloc[0])
+
+