SS 2025

2025_Research_Software_Engineering_1/src/analysis.py

@@ -0,0 +1,399 @@
+import pandas as pd
+import numpy as np
+import matplotlib.pyplot as plt
+from typing import Dict, List, Tuple
+
+DATA_PATH = "../data/13311-0001_en_flat.csv"
+
+primary_sectors: List[str] = [
+    'Agriculture, forestry and fishing',
+    'Mining and quarrying',
+    'Electricity, gas, steam, air conditioning supply',
+    'Water supply,sewerage,waste management,remediation'
+]
+
+secondary_sectors: List[str] = [
+    'Construction',
+    'Manufacturing',
+    'Industry',
+    'Industry, except construction'
+]
+
+tertiary_sectors: List[str] = [
+    'Financial, insurance, business, real estate act.',
+    'Financial and insurance activities',
+    'Real estate activities',
+    'Professional, scientific and technical activities',
+    'Administrative and support service activities',
+    'Business services',
+    'Information and communication',
+    'Wholesale, retail trade, repair of motor vehicles',
+    'Trade, transport., storage, accom.and food service',
+    'Trade,transport.,storage,accom.and food serv.,inf.',
+    'Transportation and storage',
+    'Accommodation and food service activities',
+    'Public services, education, health',
+    'Public admin. and defence, compulsory social sec.',
+    'Education',
+    'Human health and social work activities',
+    'Public a.other services, education, health,priv.h.',
+    'Other service activities',
+    'Service activities',
+    'Private households',
+    'Arts, entertainment and recreation',
+    'Arts,entertainm.,recreation,o.service act.,priv.h.'
+]
+
+
+def get_df(sector: str = 'Total',
+           etype: str = 'Persons in employment (domestic concept)',
+           start_time: int = 1999,
+           end_time: int = 2024) -> pd.DataFrame:
+    """
+    Fetches and processes a DataFrame with employment data
+    for a specified sector and employment type over a given time range.
+
+    Args:
+        sector (str): The economic sector to filter
+        data by (default is 'Total').
+        etype (str): The type of employment data to retrieve
+        (default is 'Persons in employment (domestic concept)').
+        start_time (int): The starting year of the time
+        range (default is 1999).
+        end_time (int): The ending year of the time range
+        (default is 2024).
+
+    Returns:
+        pd.DataFrame: The filtered and processed DataFrame
+        containing the employment data.
+    """
+    df: pd.DataFrame = pd.read_csv(DATA_PATH, sep=";")
+    df = df[['2_variable_attribute_label',
+             'value_variable_label',
+             'time',
+             'value']]
+    if sector != 'All':
+        df = df[df['2_variable_attribute_label'] == sector]
+    df = df[df['value_variable_label'] == etype]
+    df['time'] = pd.to_numeric(df['time'], errors='coerce')
+    df = df[(df['time'] >= start_time) & (df['time'] <= end_time)]
+    if sector == 'All':
+        df = df.sort_values(by='2_variable_attribute_label')
+    else:
+        df = df.sort_values(by='time')
+    df['value'] = pd.to_numeric(df['value'], errors='coerce')
+    if sector == 'All':
+        df = df.drop(['value_variable_label', 'time'], axis=1)
+        df = df[df['2_variable_attribute_label'] != 'Total']
+    else:
+        df = df.drop(['2_variable_attribute_label',
+                     'value_variable_label'], axis=1)
+    df = df.dropna()
+    return df
+
+
+def get_sectors() -> np.ndarray:
+    """
+    Get all unique sector names from the dataset, excluding 'Total'.
+
+    Returns:
+        np.ndarray: Array of unique sector names.
+    """
+    df: pd.DataFrame = pd.read_csv(DATA_PATH, sep=";")
+    df = df[['2_variable_attribute_label']]
+    df = df[df['2_variable_attribute_label'] != 'Total']
+    return pd.unique(df.values.ravel())
+
+
+def line_plot_df(df: pd.DataFrame, title: str) -> None:
+    """
+    Create a line plot for a single DataFrame.
+
+    Args:
+        df (pd.DataFrame): DataFrame containing 'time' and 'value' columns.
+        title (str): Title of the plot.
+
+    Returns:
+        None
+    """
+    plt.plot(df['time'], df['value'], label=title)
+    plt.title(title)
+    plt.xlabel("Time")
+    plt.ylabel("Amount")
+    plt.legend()
+    plt.show()
+
+
+def line_plot_multiple_df(labeled_dfs: Dict[str, pd.DataFrame],
+                          title: str) -> None:
+    """
+    Create a line plot for multiple DataFrames, each with its own label.
+
+    Args:
+        labeled_dfs (Dict[str, pd.DataFrame]): Dictionary where keys are
+        labels and values are DataFrames.
+        title (str): Title of the plot.
+
+    Returns:
+        None
+    """
+    for key, df in labeled_dfs.items():
+        plt.plot(df['time'], df['value'], label=key)
+    plt.title(title)
+    plt.xlabel("Time")
+    plt.ylabel("Amount")
+    plt.legend()
+    plt.show()
+
+
+def map_sector_to_group(sector: str) -> str:
+    """
+    Map a sector name to its corresponding economic
+    group (Primary, Secondary, Tertiary).
+
+    Args:
+        sector (str): The name of the sector.
+
+    Returns:
+        str: The economic group ('Primary Sector', 'Secondary Sector',
+        'Tertiary Sector', or 'Unknown').
+    """
+    if sector in primary_sectors:
+        return 'Primary Sector'
+    elif sector in secondary_sectors:
+        return 'Secondary Sector'
+    elif sector in tertiary_sectors:
+        return 'Tertiary Sector'
+    else:
+        return 'Unknown'
+
+
+def get_grouped_sector_df() -> pd.DataFrame:
+    """
+    Create a DataFrame grouping employment values by economic sector group.
+
+    Returns:
+        pd.DataFrame: DataFrame with groups ('Primary Sector', etc.)
+        and their summed employment values.
+    """
+    df: pd.DataFrame = get_df(sector='All',
+                              etype='Persons in employment (domestic concept)',
+                              start_time=1999,
+                              end_time=2024)
+    df = df.rename(columns={
+                   '2_variable_attribute_label': 'sector',
+                   'time': 'time',
+                   'value': 'value'})
+    df['Group'] = df['sector'].apply(map_sector_to_group)
+    df = df[df['Group'] != 'Unknown']
+    df_grouped = df.groupby('Group')['value'].sum().reset_index()
+    return df_grouped
+
+
+def group_values(values: List[float],
+                 labels: List[str],
+                 threshold: int = 5) -> Tuple[List[float], List[str]]:
+    """
+    Group values that are below a threshold percentage
+    into a single '<5%' category.
+
+    Args:
+        values (List[float]): List of numerical values.
+        labels (List[str]): Corresponding labels for each value.
+        threshold (int): Minimum percentage to avoid grouping. Defaults to 5%.
+
+    Returns:
+        Tuple[List[float], List[str]]: Grouped values
+        and corresponding grouped labels.
+    """
+    total: float = sum(values)
+    percentages: List[float] = [(value / total) * 100 for value in values]
+    grouped_values: List[float] = []
+    grouped_labels: List[str] = []
+    other_total: float = 0
+
+    for value, label, percent in zip(values, labels, percentages):
+        if percent < threshold:
+            other_total += value
+        else:
+            grouped_values.append(value)
+            grouped_labels.append(label)
+
+    if other_total > 0:
+        grouped_values.append(other_total)
+        grouped_labels.append("<5%")
+
+    return grouped_values, grouped_labels
+
+
+def sort_values(values: List[float],
+                labels: List[str]) -> Tuple[List[float], List[str]]:
+    """
+    Sort values and their labels in descending order based on the values.
+
+    Args:
+        values (List[float]): List of numerical values.
+        labels (List[str]): Corresponding labels for each value.
+
+    Returns:
+        Tuple[List[float], List[str]]: Sorted values and labels.
+    """
+    sorted_pairs: List[Tuple[float, str]] = sorted(
+        zip(values, labels), reverse=True)
+    sorted_values, sorted_labels = zip(*sorted_pairs)
+    return sorted_values, sorted_labels
+
+
+def pie_plot_df(values: List[float], labels: List[str], title: str) -> None:
+    """
+    Create a pie chart from values and labels, grouping small values.
+
+    Args:
+        values (List[float]): List of numerical values.
+        labels (List[str]): Corresponding labels.
+        title (str): Title of the plot.
+
+    Returns:
+        None
+    """
+    grouped_values, grouped_labels = group_values(values, labels, 5)
+    sorted_values, sorted_labels = sort_values(grouped_values, grouped_labels)
+    plt.pie(x=sorted_values, labels=sorted_labels, autopct='%1.1f%%')
+    plt.axis('equal')
+    plt.title(title)
+    plt.show()
+
+
+def pie_plot_grouped_sector(df_grouped: pd.DataFrame, title: str) -> None:
+    """
+    Create a pie chart from a grouped sector DataFrame.
+
+    Args:
+        df_grouped (pd.DataFrame): DataFrame with 'Group' and 'value' columns.
+        title (str): Title of the plot.
+
+    Returns:
+        None
+    """
+    plt.figure(figsize=(8, 8))
+    plt.pie(x=df_grouped['value'],
+            labels=df_grouped['Group'],
+            autopct='%1.1f%%',
+            startangle=140)
+    plt.title(title)
+    plt.axis('equal')
+    plt.show()
+
+
+def persons_in_employment_timeline() -> None:
+    """
+    Plot a timeline of total persons in employment from 1999 to 2024 in Germany.
+
+    Returns:
+        None
+    """
+    df_total = get_df()
+    line_plot_df(df_total,
+                 'Persons in Employment from 1999-2024 in GER')
+
+
+def employment_types_timeline() -> None:
+    """
+    Plot timelines comparing employees and self-employed persons 
+    from 1999 to 2024 in Germany.
+
+    Returns:
+        None
+    """
+    empl_attr = 'Employees (domestic concept)'
+    self_empl_attr = 'Self-employed persons a. family workers (domestic)'
+    employment_data_by_type = {
+        empl_attr: get_df(etype=empl_attr),
+        self_empl_attr: get_df(etype=self_empl_attr)
+    }
+    line_plot_multiple_df(employment_data_by_type,
+                          'Employment types from 1999-2024 in GER')
+
+
+def employed_per_sector_2024() -> None:
+    """
+    Create a pie chart showing the distribution of employees across sectors
+    in Germany for the year 2022.
+
+    Returns:
+        None
+    """
+    df_sectors_2022 = get_df(
+        sector='All',
+        etype='Employees (domestic concept)',
+        start_time=2022,
+        end_time=2022)
+    labels = df_sectors_2022['2_variable_attribute_label'].tolist()
+    pie_plot_df(values=df_sectors_2022['value'].to_list(
+    ), labels=labels, title='Most Employed per sector in 2022 in GER')
+
+
+def self_employed_per_sector_2024() -> None:
+    """
+    Create a pie chart showing the distribution of self-employed persons
+    across sectors in Germany for the year 2022.
+
+    Returns:
+        None
+    """
+    df_self_sectors_2022 = get_df(
+        sector='All',
+        etype='Self-employed persons a. family workers (domestic)',
+        start_time=2022,
+        end_time=2022)
+    labels = df_self_sectors_2022['2_variable_attribute_label'].tolist()
+    pie_plot_df(values=df_self_sectors_2022['value'].to_list(
+    ),
+        labels=labels,
+        title='Most Self-Employed per sector in 2022 in GER')
+
+
+def biggest_employment_sectors_timeline() -> None:
+    """
+    Plot timelines of employment data for the most significant sectors
+    in Germany from 1999 to 2024.
+
+    Returns:
+        None
+    """
+    dict_sectors = {
+        'Service activities': get_df(sector='Service activities'),
+        'Private households': get_df(sector='Private households'),
+        'Education': get_df(sector='Education'),
+        'Trade,transport.,storage,accom.and food serv.,inf.': get_df(
+            sector='Trade,transport.,storage,accom.and food serv.,inf.'),
+        'Business services': get_df(sector='Business services'),
+        'Public services, education, health': get_df(
+            sector='Public services, education, health'),
+        'Manufacturing': get_df(sector='Manufacturing')
+    }
+
+    emp_sector_title = 'Biggest sectors from 1999-2024 in GER'
+    line_plot_multiple_df(dict_sectors, emp_sector_title)
+
+
+def economic_sectors() -> None:
+    """
+    Create a pie chart showing the total employment distribution
+    by economic sector group (Primary, Secondary, Tertiary).
+
+    Returns:
+        None
+    """
+    df_grouped_sector = get_grouped_sector_df()
+    pie_plot_grouped_sector(
+        df_grouped_sector, 'Employment Distribution by Economic Sector')
+
+
+if __name__ == "__main__":
+    persons_in_employment_timeline()
+    employment_types_timeline()
+    employed_per_sector_2024()
+    self_employed_per_sector_2024()
+    biggest_employment_sectors_timeline()
+    economic_sectors()