Field Mapping

This page maps Odoo technical field names to the CSV columns expected by the ML training pipeline.

Revenue Model Features

Odoo Object	Odoo Field	CSV Column	Type	Notes
sale.order.line	product_uom_qty	quantity_sold	Float	Units per line item
sale.order.line	price_unit	price	Float	Price before discount
sale.order.line	discount	discount_percent	Float	0–100
sale.order	date_order	month	Int	Extracted month (1–12)
sale.order	date_order	quarter	Int	Extracted quarter (1–4)
computed	—	seasonal_index	Float	month / 6.5 (normalised)
computed	—	gross_margin_pct	Float	(price - cost) / price
sale.order.line	price_subtotal	net_revenue	Float	Target variable

Churn Model Features

Odoo Object	Odoo Field	CSV Column	Type	Notes
res.partner	aggregated	days_since_last_order	Int	Days since last sale.order
res.partner	aggregated	total_orders	Int	COUNT of confirmed orders
res.partner	aggregated	avg_days_between_orders	Float	Mean inter-order gap
res.partner	custom field	churn_risk_score	Float	0–1 CRM risk score
res.partner	aggregated	total_spent	Float	SUM of amount_total
res.partner	aggregated	avg_order_value	Float	AVG of amount_total
res.partner	computed	churned	Int	Target (0 = active, 1 = churned)

Growth Model Features

Source	CSV Column	Type	Notes
Sales analysis	period	String	e.g. "2026-Q1"
Sales analysis	total_revenue	Float	Period revenue total
Sales analysis	order_count	Int	Orders in period
Sales analysis	new_customers	Int	New partners in period
Sales analysis	avg_order_value	Float	Period average
Computed (3m lag)	revenue_growth_3m	Float	Target variable

Computed Field Reference

seasonal_index

seasonal_index = month / 6.5
# Ranges approximately 0.15 (Jan) → 1.85 (Dec)
# Captures cyclical monthly patterns

churned Label

# A customer is churned if they have not ordered in 90+ days
from datetime import datetime, timedelta
cutoff = datetime.now() - timedelta(days=90)
df['churned'] = (df['last_order_date'] < cutoff).astype(int)

revenue_growth_3m (target)

# 3-period forward revenue (shifted for supervised learning)
df['revenue_growth_3m'] = df['total_revenue'].shift(-3)
df.dropna(subset=['revenue_growth_3m'], inplace=True)