Added processing to apply project credits, determine institution name…

… for each PI, and exporting HU and BU invoices

process_report/process_report.py

@@ -1,9 +1,105 @@
 import argparse
 import os
+import sys
 
 import pandas
 
 
+### Invoice field names
+PROJECT_FIELD = 'Project - Allocation'
+MANAGER_PI_FIELD = 'Manager (PI)'
+INSTITUTION_FIELD = 'Institution'
+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()
+    try:
+        with open(os.getenv("C0002_OLD_PI")) as f:
+            for pi in f: old_pi_list.add(pi.strip())
+
+    except Exception:
+        print("Old PI file does not exist, or the C0002_OLD_PI env var is not set")
+        sys.exit(1)
+
+    pi_list = dataframe[MANAGER_PI_FIELD].unique()
+
+    for pi in pi_list:
+        if pi != pi: continue   # NaN check
+        if pi in old_pi_list: continue  # Is the PI an old PI?
+
+        pi_projects = dataframe[dataframe[MANAGER_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):
+
+    institute_map = {
+        "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",
+    }
+
+    for name, institute in institute_map.items():
+        if name in pi_uname: return institute
+
+    print(f"PI name {pi_uname} does not match any institution!")
+    return ""
+
+
 def main():
     """Remove non-billable PIs and projects"""
 
@@ -41,6 +137,18 @@ 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 +168,13 @@ 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):
@@ -102,7 +214,7 @@ 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[MANAGER_PI_FIELD].isin(pi) & ~dataframe[PROJECT_FIELD].isin(projects)]
     filtered_dataframe.to_csv(output_file, index=False)
     return filtered_dataframe
 
@@ -112,21 +224,57 @@ 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[MANAGER_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()
+    pi_list = dataframe[MANAGER_PI_FIELD].unique()
 
     for pi in pi_list:
-        pi_projects = dataframe[dataframe['Manager (PI)'] == pi]
-        pi_instituition = pi_projects['Institution'].iat[0]
+        if pi != pi: continue
+        pi_projects = dataframe[dataframe[MANAGER_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[MANAGER_PI_FIELD]
+        if pi_name != 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()