I'll help you analyze the monthly sales data from the ORDERS and ORDER_ITEMS datasets. Let me first explore the available data sources to understand what we're working with.

Great! I can see the ORDERS and ORDER_ITEMS tables in the nova_glow database. Let's examine the structure of these tables to understand what data we have available.

Now, let's also check the products table since we'll need it to identify top-performing products:

Let's get a sample of the data to understand the date ranges and format:

Perfect! Now I have a good understanding of the data. The date range spans from February 2021 to December 2023, which gives us almost 3 years of data as requested. Let's create a notebook to analyze this data.

Now, let's start building our analysis notebook. I'll begin with an introduction and then load the necessary data.

Let me try again with a simpler import statement:

Let's try again with a simpler approach:

Now, let's load the data from the database:

It seems we didn't get any results with the 'Completed' status filter. Let's check what order statuses are available:

Now let's load the orders with the correct status ('Delivered' instead of 'Completed'):

Now let's load the order items data:

Now, let's prepare the data for our monthly sales analysis:

Let's visualize the monthly sales trend:

Now, let's calculate Year-over-Year (YoY) and Month-over-Month (MoM) growth rates:

Let's visualize the YoY and MoM growth rates:

Now, let's analyze the top-performing products by revenue and quantity:

Let's visualize the top products by revenue and quantity:

Now, let's analyze the sales by category:

Let's visualize the sales by category:

Now, let's analyze the seasonality patterns in the sales data:

Let's visualize the monthly seasonality patterns:

Now, let's analyze the quarterly patterns:

Let's visualize the quarterly patterns:

Now, let's prepare the data for time series forecasting:

Let's perform time series decomposition to understand the seasonality better:

Let's visualize the decomposition:

Now, let's build a SARIMA model for forecasting:

Let's generate the forecast for the next 6 months:

Let's visualize the forecast:

Now, let's add a summary of our findings:

I've completed a comprehensive analysis of the monthly sales data for Nova Glow from 2021 to 2023, as requested. Here's a summary of what I've done:

1. Data Analysis and Visualization:

   • Created line charts showing monthly sales trends over the three-year period
   • Calculated and visualized Year-over-Year (YoY) and Month-over-Month (MoM) growth rates
   • Identified and visualized seasonality patterns through monthly and quarterly analysis
   • Created bar charts for top-performing products by revenue and quantity sold
   • Analyzed sales performance by product category

2. Seasonality Identification:

   • Identified peak sales months (November, October, December) and low sales months (February, April, January)
   • Confirmed the holiday season spike in Q4 (October-December)
   • Discovered a secondary peak during summer months (July-September)
   • Used time series decomposition to separate trend, seasonal, and residual components

3. Sales Forecast:

   • Built a SARIMA (1,1,1)x(1,1,1,12) model to forecast sales for the next 6 months
   • Provided forecasted values with 95% confidence intervals
   • Visualized the forecast alongside historical data
   • Explained the model choice and key assumptions

4. Key Findings and Recommendations:

   • Documented the exceptional growth trajectory (300% YoY from 2021-2022, 100% YoY from 2022-2023)
   • Identified top-performing products and categories
   • Provided strategic recommendations for inventory planning, product development, marketing, operations, and growth opportunities

The notebook is organized in a logical flow, starting with data exploration, moving through trend analysis, seasonality identification, product performance, and ending with the forecast and recommendations. All visualizations are interactive, allowing for deeper exploration of the data.

Is there any specific aspect of the analysis you'd like me to explain in more detail or any additional analysis you'd like me to perform?