import warnings; warnings.filterwarnings("ignore")
# Import pandas and matplotlib for data handling and plotting
import pandas as pd
import matplotlib.pyplot as plt

# Set up plots to display in the notebook
%matplotlib inline

# Load shampoo sales time series from a CSV online
url = "https://raw.githubusercontent.com/jbrownlee/Datasets/master/shampoo.csv"
df = pd.read_csv(url)

# Show data size and first five rows
print("Data shape:", df.shape)
df.head()

# Rename columns for easier access
df.columns = ["Month", "Sales"]
# Convert Month to datetime so pandas understands dates
df["Month"] = pd.date_range(start="1901-01", periods=len(df), freq="M")
df["Month"] = pd.to_datetime(df["Month"])

df.head()

# Plot shampoo sales over time
plt.figure(figsize=(10,5))
plt.plot(df["Month"], df["Sales"], marker="o")
plt.xlabel("Month")
plt.ylabel("Sales")
plt.title("Monthly Shampoo Sales")
plt.grid(True)
plt.tight_layout()
plt.show()

# Check for missing sales values
missing = df["Sales"].isnull().sum()
print("Missing sales values:", missing)

# Fill missing values with the previous month's sales
df["Sales"] = df["Sales"].fillna(method="ffill")
df.head()

# Find months with highest and lowest sales
max_idx = df["Sales"].idxmax()
min_idx = df["Sales"].idxmin()
max_month = df.loc[max_idx, "Month"]
min_month = df.loc[min_idx, "Month"]

print("Peak sales in:", max_month.strftime("%Y-%m"))
print("Low sales in:", min_month.strftime("%Y-%m"))

# Add peak and low points to the chart
plt.figure(figsize=(10,5))
plt.plot(df["Month"], df["Sales"], label="Monthly Sales", marker="o")
plt.scatter(df.loc[max_idx, "Month"], df.loc[max_idx, "Sales"], color="red", label="Peak")
plt.scatter(df.loc[min_idx, "Month"], df.loc[min_idx, "Sales"], color="blue", label="Low")
plt.xlabel("Month")
plt.ylabel("Sales")
plt.title("Shampoo Sales with Peaks and Lows")
plt.legend()
plt.grid(True)
plt.tight_layout()
plt.show()

# Change style and add markers to plot
plt.style.use("seaborn-v0_8-colorblind")
plt.figure(figsize=(10,5))
plt.plot(df["Month"], df["Sales"], marker="D", linestyle="--", color="purple")
plt.xlabel("Month")
plt.ylabel("Sales (units)")
plt.title("Shampoo Sales: Styled Plot")
plt.grid(True)
plt.tight_layout()
plt.show()

# Zoom in on just the first year of sales
first_year = df[df["Month"].dt.year == 1901]
plt.figure(figsize=(8,4))
plt.plot(first_year["Month"], first_year["Sales"], marker="o", color="green")
plt.title("Shampoo Sales in 1901")
plt.xlabel("Month")
plt.ylabel("Sales")
plt.tight_layout()
plt.show()

# Ask the user to choose a year and plot it
year = int(input("Enter a year between 1901 and 1903: "))
year_data = df[df["Month"].dt.year == year]
plt.figure(figsize=(8,4))
plt.plot(year_data["Month"], year_data["Sales"], marker="s", color="orange")
plt.title(f"Shampoo Sales in {year}")
plt.xlabel("Month")
plt.ylabel("Sales")
plt.tight_layout()
plt.show()

# Running statistics to see trends
df["Rolling Mean"] = df["Sales"].rolling(window=3).mean()
plt.figure(figsize=(10,5))
plt.plot(df["Month"], df["Sales"], label="Monthly Sales")
plt.plot(df["Month"], df["Rolling Mean"], label="3-Month Average", color="red")
plt.xlabel("Month")
plt.ylabel("Sales")
plt.title("Sales with 3-Month Moving Average")
plt.legend()
plt.grid(True)
plt.tight_layout()
plt.show()

# Filter months with sales above 400
high_sales = df[df["Sales"] > 400]
print("Months with sales above 400:")
print(high_sales[["Month", "Sales"]])

# Sort dataset by sales from lowest to highest
sorted_df = df.sort_values("Sales")
print(sorted_df[["Month", "Sales"]].head())

# Calculate monthly sales increase
df["Change"] = df["Sales"].diff()
print(df[["Month", "Sales", "Change"]].head(10))

# Plot all years as separate lines for easy comparison
plt.figure(figsize=(10,6))
for year, group in df.groupby(df["Month"].dt.year):
    plt.plot(group["Month"], group["Sales"], marker="o", label=str(year))
plt.xlabel("Month")
plt.ylabel("Sales")
plt.title("Shampoo Sales by Year")
plt.legend(title="Year")
plt.grid(True)
plt.tight_layout()
plt.show()

# Plot all years and mark biggest jump and sharpest drop
plt.figure(figsize=(12,6))
for year, group in df.groupby(df["Month"].dt.year):
    plt.plot(group["Month"], group["Sales"], marker="o", label=str(year))
jump_idx = df["Change"].idxmax()
drop_idx = df["Change"].idxmin()
plt.scatter(df.loc[jump_idx, "Month"], df.loc[jump_idx, "Sales"],
            color="green", s=100, label="Biggest Jump")
plt.scatter(df.loc[drop_idx, "Month"], df.loc[drop_idx, "Sales"],
            color="red", s=100, label="Sharpest Drop")
plt.xlabel("Month")
plt.ylabel("Sales")
plt.title("Market Report: Sales Jumps and Drops")
plt.legend()
plt.grid(True)
plt.tight_layout()
plt.show()

# Save your latest plot to a file
plt.figure(figsize=(12,6))
for year, group in df.groupby(df["Month"].dt.year):
    plt.plot(group["Month"], group["Sales"], marker="o", label=str(year))
plt.xlabel("Month")
plt.ylabel("Sales")
plt.title("Shampoo Sales by Year")
plt.legend()
plt.grid(True)
plt.tight_layout()
plt.savefig("shampoo_sales_by_year.png")
plt.close()
print("Plot saved as shampoo_sales_by_year.png")

# Basic troubleshooting: Check for bad data types
print(df.dtypes)