# Step 1: Python setup and warning suppression
import warnings
warnings.filterwarnings("ignore")
print("Hello! Welcome to your time series adventure.")

# Data setup: Let us load daily minimum temperature data
import pandas as pd
url = "https://raw.githubusercontent.com/jbrownlee/Datasets/master/daily-min-temperatures.csv"
data = pd.read_csv(url)
print("Data shape:", data.shape)
print("First 5 rows:")
print(data.head())

# Plot the temperature time series
import matplotlib.pyplot as plt
plt.figure(figsize=(10,4))
plt.plot(data["Temp"])
plt.title("Daily Minimum Temperatures")
plt.xlabel("Day")
plt.ylabel("Temperature (C)")
plt.grid(True)
plt.show()

# Calculate and plot a rolling mean (trend line)
data["Trend"] = data["Temp"].rolling(window=365).mean()
plt.figure(figsize=(10,4))
plt.plot(data["Temp"], label="Daily")
plt.plot(data["Trend"], color="red", linewidth=3, label="Trend (365-day avg)")
plt.title("Trend in Daily Temperatures")
plt.legend()
plt.show()

# Highlighting seasonality with monthly averages
data["Month"] = pd.to_datetime(data["Date"]).dt.month
monthly_means = data.groupby("Month")["Temp"].mean()
plt.figure(figsize=(8,4))
plt.bar(monthly_means.index, monthly_means.values)
plt.title("Average Temperature by Month")
plt.xlabel("Month (1=Jan, 12=Dec)")
plt.ylabel("Average Temp (C)")
plt.show()

# Use smoothing to try spotting cycles (longer rolling window)
data["Cyclic"] = data["Temp"].rolling(window=730).mean()  # about 2 years
plt.figure(figsize=(10,4))
plt.plot(data["Temp"], alpha=0.3, label="Daily Temps")
plt.plot(data["Cyclic"], color="purple", linewidth=3, label="Cyclic (2yr avg)")
plt.title("Possible Cyclic Patterns in Temperature")
plt.legend()
plt.show()

# Visualizing noise: subtract trend and seasonality
noisy_values = data["Temp"] - data["Trend"]
plt.figure(figsize=(10,4))
plt.plot(noisy_values, color="gray")
plt.title("Just the Noise (original minus trend)")
plt.xlabel("Day")
plt.ylabel("Noise")
plt.show()

# Error handling: what if a column is missing?
try:
    data["FakeCol"]
except KeyError:
    print("Column not found! Check your spelling or column names.")
    

# Updating data with new columns: flag cold days
data["ColdDay"] = data["Temp"] < 5
print(data[["Date", "Temp", "ColdDay"]].head())

# Remove unwanted columns safely
if "FakeCol" in data.columns:
    data = data.drop(columns=["FakeCol"])
print("Columns now:", data.columns.tolist())

# Useful: Count number of cold days
num_cold = data["ColdDay"].sum()
print(f"Number of cold days: {num_cold}")

# Loop: print the first 5 cold days
count = 0
for idx, row in data.iterrows():
    if row["ColdDay"]:
        print(row["Date"], row["Temp"])
        count += 1
        if count == 5:
            break
        

# List comprehensions: list years with more than 60 cold days
data["Year"] = pd.to_datetime(data["Date"]).dt.year
years = data["Year"].unique()
cold_years = [year for year in years if data[data["Year"] == year]["ColdDay"].sum() > 60]
print(cold_years)

# Sorting: hottest days in the dataset
hottest = data.sort_values("Temp", ascending=False).head(5)
print(hottest[["Date","Temp"]])

# Filtering: only show summer months (Dec-Jan-Feb in Australia)
summer = data[data["Month"].isin([12,1,2])]
print(summer[["Date","Temp"]].head())

# Combining data: average summer vs. winter temperature
winter = data[data["Month"].isin([6,7,8])]
avg_summer = summer["Temp"].mean()
avg_winter = winter["Temp"].mean()
print(f"Avg summer temp: {avg_summer:.1f} C, Avg winter temp: {avg_winter:.1f} C")

# Mini-project: ask user for a temperature cutoff and count
cut = input("Enter a temperature cutoff (Celsius): ")
cut = float(cut)
below = data[data["Temp"] < cut]
print(f"Number of days below {cut}C: {len(below)}")

# Mini-project, part 2: plot those cold days you chose
plt.figure(figsize=(10,4))
plt.plot(data["Temp"], label="All temps")
plt.scatter(below.index, below["Temp"], color="red", label="Below cutoff")
plt.axhline(y=cut, color="red", linestyle="--")
plt.title(f"Days Below {cut} Celsius")
plt.legend()
plt.show()

# Troubleshooting: handle missing values
missing = data["Temp"].isnull().sum()
print("Missing temperature values:", missing)
if missing > 0:
    data["Temp"] = data["Temp"].fillna(data["Temp"].mean())
    print("Filled missing values with the mean.")
    

# Quick tip: see which columns are numeric
numeric_cols = data.select_dtypes(include="number").columns.tolist()
print(numeric_cols)

# Challenge: ask for a month number, count days above 20C
m = input("Pick a month number (1-12): ")
m = int(m)
in_month = data[data["Month"] == m]
above_20 = in_month[in_month["Temp"] > 20]
print(f"Days above 20C in month {m}: {len(above_20)}")