analyze script is included to repo

utils/analyze.py

@@ -0,0 +1,87 @@
+import pandas as pd
+import matplotlib.pyplot as plt
+from pandas.core.internals.blocks import F
+# example run on Google colab
+# https://colab.research.google.com/drive/17Zn2y9QnjWchwNdBpJ4brjj7IPzH5I_5?usp=sharing
+
+KPI_COLUMNS = [
+    "Operating revenue (Turnover)\nm USD",
+    "Sales\nm USD",
+    "Gross profit\nm USD",
+    "P/L before tax\nm USD",
+    "P/L for period [=Net income]\nm USD",
+    "Cash flow\nm USD",
+    "Total assets\nm USD",
+    # "Number of employees\n",
+]
+available_years = [str(year) for year in range(1994, 2024)]
+
+
+def create_data_dictionary(df):
+    company_missing_kpi_mapping = {}
+    # # Iterate over each row
+    for index, row in df.iterrows():
+        agreement_date_year = str(row["agreementDate"]).split("-")[0]
+        if agreement_date_year != "nan":
+            agreement_date_year = str(int(agreement_date_year) - 1)
+            company_name = row["Company name Latin alphabet_x"]
+            if company_name not in d:
+                company_missing_kpi_mapping[company_name] = {}
+            else:
+                if "missing_kpi" not in company_missing_kpi_mapping[company_name]:
+                    company_missing_kpi_mapping[company_name]["missing_kpi"] = 0
+                else:
+                    for column in KPI_COLUMNS:
+                        if int(agreement_date_year) < 1994:
+                            company_missing_kpi_mapping[company_name][f"{column} {agreement_date_year}"] = False
+                        else:
+                            print(row[f"{column} {agreement_date_year}"])
+                            if (
+                                row[f"{column} {agreement_date_year}"] == "n.a."
+                                or row[f"{column} {agreement_date_year}"] is None
+                            ):
+                                company_missing_kpi_mapping[company_name]["missing_kpi"] += 1
+                                company_missing_kpi_mapping[company_name][f"{column} {agreement_date_year}"] = False
+    return company_missing_kpi_mapping
+
+
+def create_missing_kpi_var_map(data: dict, is_licensee: bool = True):
+    # Extracting companies with missing_kpi values
+    companies = [company for company in data.keys() if data[company].get("missing_kpi", 0) > 0]
+    missing_kpis = [data[company]["missing_kpi"] for company in companies]
+
+    # Sorting companies based on missing_kpi values (high to low)
+    sorted_companies, sorted_missing_kpis = zip(*sorted(zip(companies, missing_kpis), key=lambda x: x[1], reverse=True))
+
+    # Prepare the total number of missing KPI variables for each company
+    missing_kpi_vars = []
+    for company in sorted_companies:
+        missing_vars = [var for var in data[company] if data[company][var] is False]
+        missing_kpi_vars.append(", ".join(missing_vars))
+
+    # Plotting the bar chart
+    plt.figure(figsize=(12, 6))
+    bars = plt.bar(range(len(sorted_companies)), sorted_missing_kpis)
+    plt.xlabel('Company')
+    plt.ylabel('Number of Missing KPIs')
+    if is_licensee:
+        plt.title('Companies with Missing KPIs on merged data (Licensee)')
+    else:
+        plt.title('Companies with Missing KPIs on merged data (Licensor)')
+    plt.xticks(range(len(sorted_companies)), sorted_companies, rotation=90)
+
+    # Adding values inside the bars
+    for bar, value in zip(bars, sorted_missing_kpis):
+        plt.text(bar.get_x() + bar.get_width() / 2, bar.get_height() - 1, str(value), ha='center', va='center', fontsize=10, color='white')
+
+    # Adding total number of missing KPI variables as text in the right upper corner
+    total_missing_kpi_vars = sum(len(missing_vars.split(", ")) for missing_vars in missing_kpi_vars)
+    table_total_text = f"Total Missing KPI Variables: {total_missing_kpi_vars}"
+    plt.text(0.65, 0.90, table_total_text, transform=plt.gca().transAxes, fontsize=12, fontweight='bold')
+
+    plt.tight_layout()
+    plt.show()
+
+
+
+