Build a sales dashboard from one cleaned table

This example cleans a raw sales export once, then forks into three views: a product leaderboard measured against per-line revenue targets, a revenue matrix by city and customer type, and a monthly revenue trend. It's for analysts who have outgrown a single summary and want several angles on the same data without repeating the cleaning each time. It builds on the five-node sales pipeline; read that one first if branching flows are new to you.

Flow: sales_dashboard.yaml · In-app: Create → From template → "Sales dashboard: rank, pivot, and trend" · Data: data/templates/supermarket_sales.csv

See it: the finished dashboard on the canvas

The data

supermarket_sales.csv holds 1030 transaction rows for the 2024 calendar year — including 30 exact duplicate rows to clean up.

Column	Meaning
invoice_id	Transaction identifier
city	Store city (Bago, Mandalay, Naypyitaw, Taunggyi, Yangon)
customer_type	Member or Normal
product_line	Product category (six of them)
unit_price	Price per unit
quantity	Units sold
gross_income	Income for the line
date	Transaction date

The flow

The flow shares one cleaned, enriched stream and branches it three ways.

Prepare the data (the trunk):

1. Read data — reads supermarket_sales.csv (file type CSV). The date column is parsed as a real date.
2. Drop duplicates — Unique with strategy any and no key columns, removing the 30 fully-identical rows.
3. Formula — adds revenue = [unit_price] * [quantity]. This node has three outgoing connections; each branch below reads from it.

See it: the enriched trunk and its three-way fork

Branch 1 — product leaderboard vs target:

1. Group by — groups by product_line, aggregating revenue (sum → total_revenue), invoice_id (n-unique → orders), and quantity (sum → units_sold).
2. Manual input — a six-row lookup giving each product_line a department and an annual_target.
3. Join — a left join of the leaderboard (main input) to the lookup (right input) on product_line, carrying department and annual_target onto each row.
4. Formula — adds pct_of_target = [total_revenue] / [annual_target] * 100.
5. Sort — total_revenue descending, so the best-selling line is on top.
6. Explore data — opens the ranking in the Graphic Walker explorer.

See it: the leaderboard joined to its targets

Branch 2 — revenue by city and customer type:

1. Pivot — index city, pivot column customer_type, values revenue, aggregation sum. This reshapes the data into one row per city with a Member and a Normal column.
2. Explore data — opens the matrix.

See it: the city × customer-type pivot

Branch 3 — monthly trend:

1. Formula — adds month = format_date([date], "%Y-%m"), turning each date into a 2024-01-style label.
2. Group by — groups by month, summing revenue → monthly_revenue.
3. Sort — month ascending, so the months read in order.
4. Explore data — opens the trend.

See it: the monthly trend branch

Run it

• In your browser — open it in Flowfile Lite: the flow travels in the link and reads the sample CSV from its public URL. Click Run when it loads.
• From the template browser — Create → From template → "Sales dashboard: rank, pivot, and trend". The sample data is wired in; click Run.
• Download and open — grab sales_dashboard.yaml and open it in the designer. Its Read node points at the sample CSV's public URL, so it runs as-is with an internet connection — or repoint it at a local copy.
• Headless — once saved with a real data path, run it from the command line:

flowfile run flow path/to/your_flow.yaml

The result

Product leaderboard vs target (Branch 1), over the deduplicated rows:

product_line	total_revenue	orders	units_sold	department	annual_target	pct_of_target
Home and lifestyle	54383.40	182	1030	Homeware	55000	98.88
Fashion accessories	53623.70	167	958	Apparel	50000	107.25
Food and beverages	53446.94	169	980	Grocery	52000	102.78
Sports and travel	51602.13	169	977	Leisure	50000	103.20
Health and beauty	47912.69	158	878	Personal care	48000	99.82
Electronic accessories	46463.40	155	869	Electronics	45000	103.25

See it: the leaderboard charted in Graphic Walker

Revenue by city and customer type (Branch 2):

city	Member	Normal
Bago	30431.18	28633.08
Mandalay	37913.27	30676.55
Naypyitaw	26816.89	21739.14
Taunggyi	31751.35	35469.53
Yangon	30491.67	33509.60

See it: the revenue matrix charted

Monthly trend (Branch 3) runs from 2024-01 at 29817.92 to 2024-12 at 29917.37, one row per month.

See it: the monthly trend line

In Python

The same three views with the FlowFrame API — clean once into base, then derive each view from it. Every source is built into one graph (note flow_graph=graph on the read and the lookup), so ff.open_graph_in_editor(graph) opens all three branches back in the visual editor:

import flowfile as ff

SALES = "https://raw.githubusercontent.com/edwardvaneechoud/flowfile/main/data/templates/supermarket_sales.csv"

# One graph holds every node, so all three views land in a single flow you can open.
graph = ff.create_flow_graph()

# Clean and enrich once — every view below reads from this.
base = (
    ff.read_csv(SALES, flow_graph=graph)
    .unique()
    .with_columns((ff.col("unit_price") * ff.col("quantity")).alias("revenue"))
)

# A small lookup: each product line's department and annual revenue target.
# Passing flow_graph=graph keeps it on the same graph, so the join below doesn't fork a new one.
targets = ff.from_dict(
    {
        "product_line": [
            "Home and lifestyle", "Fashion accessories", "Food and beverages",
            "Sports and travel", "Health and beauty", "Electronic accessories",
        ],
        "department": ["Homeware", "Apparel", "Grocery", "Leisure", "Personal care", "Electronics"],
        "annual_target": [55000, 50000, 52000, 50000, 48000, 45000],
    },
    flow_graph=graph,
)

# View 1 — product leaderboard, joined to targets and ranked by revenue.
products = (
    base.group_by("product_line")
    .agg(
        ff.col("revenue").sum().alias("total_revenue"),
        ff.col("invoice_id").n_unique().alias("orders"),
        ff.col("quantity").sum().alias("units_sold"),
    )
    .join(targets, on="product_line", how="left")
    .with_columns((ff.col("total_revenue") / ff.col("annual_target") * 100).alias("pct_of_target"))
    .sort("total_revenue", descending=True)
)

# View 2 — revenue matrix: one row per city, one column per customer type.
city_matrix = base.pivot(on="customer_type", index="city", values="revenue", aggregate_function="sum")

# View 3 — monthly revenue trend.
monthly = (
    base.with_columns(ff.col("date").dt.strftime(format="%Y-%m").alias("month"))
    .group_by("month")
    .agg(ff.col("revenue").sum().alias("monthly_revenue"))
    .sort("month")
)

# All three branches live in `graph` — open the whole flow in the visual editor:
# ff.open_graph_in_editor(graph)

Variations

• Pull targets from elsewhere — the targets live inline in a Manual Input here; replace it with a Read node (a spreadsheet, or a database table) to keep the Join wired to real, maintained targets.
• Change the pivot axes — swap customer_type for product_line on the Pivot node to see each product's revenue per city.
• Trend a different measure — point Branch 3's Group by at gross_income or quantity instead of revenue.
• Persist a view — replace any Explore data node with a Catalog Writer to save that view as a catalog table, then chart it with a visualization.