import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns

OUT = "eda_outputs"
import os
os.makedirs(OUT, exist_ok=True)

df = pd.read_csv("PaavoDataset.csv", encoding="latin1", sep=";")
df.columns = df.columns.str.strip()  # remove stray leading/trailing spaces
df = df[~df["Area"].str.strip().str.startswith("MK")].copy() # Drop regional summary rows (e.g. "MK01 Uusimaa")

#Summary stats
print("--- Info ---")
print(df.info())
print("\n--- Descriptive statistics ---")
desc = df.describe().T
print(desc)
desc.to_csv(f"{OUT}/summary_statistics.csv")

# Distributions of key variables
key_vars = [
    "Inhabitants",
    "Average age of inhabitants",
    "Average income of inhabitants",
    "Median income of inhabitants",
    "Average size of households",
]

fig, axes = plt.subplots(len(key_vars), 1, figsize=(8, 4 * len(key_vars)))
for ax, col in zip(axes, key_vars):
    sns.histplot(df[col], kde=True, ax=ax, color="#6fa8dc")
    ax.set_title(f"Distribution of {col}")
plt.tight_layout()
fig.savefig(f"{OUT}/01_distributions.png", dpi=150)
plt.close(fig)

# Correlation heatmap (numeric columns only)
numeric_df = df.select_dtypes(include="number")
corr = numeric_df.corr()

fig, ax = plt.subplots(figsize=(14, 12))
sns.heatmap(corr, cmap="coolwarm", center=0, annot=False, ax=ax)
ax.set_title("Correlation heatmap")
fig.savefig(f"{OUT}/02_correlation_heatmap.png", dpi=150, bbox_inches="tight")
plt.close(fig)

# Top areas by income and population
top_income = df.nlargest(15, "Average income of inhabitants")[["Area", "Average income of inhabitants"]]
top_population = df.nlargest(15, "Inhabitants")[["Area", "Inhabitants"]]
print("\n--- Top 15 areas by average income ---")
print(top_income)
print("\n--- Top 15 areas by population ---")
print(top_population)

# Income category composition (stacked bar, top 20 by inhabitants)
income_cats = [
    "Inhabitants belonging to the lowest income category",
    "Inhabitants belonging to the middle income category",
    "Inhabitants belonging to the highest income category",
]

top20_income = df.nlargest(20, "Inhabitants").set_index("Area")[income_cats].iloc[::-1]
top20_income.columns = ["Lowest income cat.", "Middle income cat.", "Highest income cat."]

fig, ax = plt.subplots(figsize=(10, 8))
top20_income.plot(kind="barh", stacked=True, color=["#6fa8dc", "#cc6659", "#e69138"], ax=ax)
ax.set_title("Income category composition - top 20 areas by population")
ax.set_xlabel("Number of inhabitants")
fig.savefig(f"{OUT}/03_income_categories_top20.png", dpi=150, bbox_inches="tight")
plt.close(fig)

# Household income category composition (same style, top 20 by households)
household_cats = [
    "Households belonging to the lowest income category",
    "Households belonging to the middle income category",
    "Households belonging to the highest income category",
]

top20_households = (df.nlargest(20, "Households").set_index("Area")[household_cats].iloc[::-1])
top20_households.columns = ["Lowest income cat.", "Middle income cat.", "Highest income cat."]

fig, ax = plt.subplots(figsize=(10, 8))
top20_households.plot(kind="barh", stacked=True, color=["#6fa8dc", "#cc6659", "#e69138"], ax=ax)
ax.set_title("Household income category composition - top 20 areas by number of households")
ax.set_xlabel("Number of households")
fig.savefig(f"{OUT}/04_household_income_categories_top20.png", dpi=150, bbox_inches="tight")
plt.close(fig)
 
# Inhabitants per residential building (housing density proxy)
df["Inhabitants per residential building"] = (
    df["Inhabitants"] / df["Residential buildings"].replace(0, pd.NA)
)
 
density_top20 = df.nlargest(20, "Inhabitants per residential building")[
    ["Area", "Inhabitants per residential building"]
].set_index("Area")
 
fig, ax = plt.subplots(figsize=(9, 8))
density_top20.sort_values("Inhabitants per residential building").plot(
    kind="barh", legend=False, color="#6fa8dc", ax=ax
)
ax.set_title("Top 20 areas by inhabitants per residential building)")
ax.set_xlabel("Inhabitants per residential building")
fig.savefig(f"{OUT}/05_inhabitants_per_building_top20.png", dpi=150, bbox_inches="tight")
plt.close(fig)
 
#  Accumulated purchasing power of inhabitants
purchasing_top20 = df.nlargest(20, "Accumulated purchasing power of inhabitants")[
    ["Area", "Accumulated purchasing power of inhabitants"]
].set_index("Area")
 
fig, ax = plt.subplots(figsize=(9, 8))
purchasing_top20.sort_values("Accumulated purchasing power of inhabitants").plot(
    kind="barh", legend=False, color="#e69138", ax=ax
)
ax.set_title("Top 20 areas by accumulated purchasing power of inhabitants")
ax.set_xlabel("Accumulated purchasing power (EUR)")
fig.savefig(f"{OUT}/06_purchasing_power_top20.png", dpi=150, bbox_inches="tight")
plt.close(fig)
 
print(f"\nAll plots and summary CSV saved to ./{OUT}/")