Power BI or Pandas First — What Beginners Should Pick

Power BI — Core Concepts Beginners Learn:

Data Sources:
  - Excel, CSV, SQL Server, SharePoint
  - Cloud apps (Salesforce, Google Analytics)
  - APIs

Power Query (ETL):
  - Load and clean data before modelling
  - Merge, append, pivot, unpivot
  - Column transformations

Data Model:
  - Star schema (fact + dimension tables)
  - Relationships (one-to-many, many-to-many)
  - Measures vs calculated columns

DAX (formulas):
  - SUM, AVERAGE, COUNT
  - CALCULATE, FILTER
  - Time intelligence (YTD, MoM)

Visualisations:
  - Bar, line, pie, map, table, matrix
  - KPI cards and gauges
  - Slicers and drill-through

Publishing:
  - Power BI Service (cloud)
  - Scheduled refresh
  - Row-level security

Power BI — Sample DAX Formulas:

Total Sales:
  Total Sales = SUM(Sales[Amount])

Average Order Value:
  AOV = DIVIDE([Total Sales], COUNTROWS(Sales))

Year-to-Date Sales:
  YTD Sales =
    CALCULATE(
      [Total Sales],
      DATESYTD('Date'[Date])
    )

Sales Last Year:
  Sales LY =
    CALCULATE(
      [Total Sales],
      SAMEPERIODLASTYEAR('Date'[Date])
    )

Year-over-Year Growth %:
  YoY % = DIVIDE([Total Sales] - [Sales LY], [Sales LY])

High-Value Customers:
  High Value = FILTER(
    Customers,
    [Total Sales] > 100000
  )

DAX is what separates basic Power BI users
from highly-paid BI developers.

# Pandas — basic operations

import pandas as pd

# Load data
df = pd.read_csv("sales.csv")

# Preview
print(df.head())
print(df.info())
print(df.describe())

# Select columns and filter
sales = df[["region", "product", "amount"]]
high  = df[df["amount"] > 1000]

# Group and aggregate
summary = df.groupby("region")["amount"].sum()

# Sort
top = df.sort_values("amount", ascending=False).head(10)

# Add a new column
df["tax"] = df["amount"] * 0.18

# Save
df.to_csv("clean_sales.csv", index=False)

# Pandas — data cleaning

import pandas as pd
import numpy as np

df = pd.read_csv("messy_sales.csv")

# Drop duplicates
df = df.drop_duplicates()

# Handle missing values
df["amount"] = df["amount"].fillna(0)
df = df.dropna(subset=["customer_id"])

# Fix data types
df["date"] = pd.to_datetime(df["date"])
df["amount"] = pd.to_numeric(df["amount"], errors="coerce")

# Clean text
df["customer"] = df["customer"].str.strip().str.title()

# Rename columns
df = df.rename(columns={"cust_id": "customer_id"})

# Handle outliers
q99 = df["amount"].quantile(0.99)
df = df[df["amount"] <= q99]

# Feature engineering
df["month"]   = df["date"].dt.month
df["quarter"] = df["date"].dt.quarter

# Ready for analysis or Power BI!

Power BI Path — 6 weeks (BI Analyst focus):

Week 1: Basics
  ☐ Install Power BI Desktop (free)
  ☐ Load Excel / CSV data
  ☐ Basic visuals (bar, line, card)

Week 2: Power Query (ETL)
  ☐ Transform data before loading
  ☐ Merge and append tables
  ☐ Pivot and unpivot

Week 3: Data Modelling
  ☐ Star schema basics
  ☐ Relationships between tables
  ☐ Calculated columns vs measures

Week 4: DAX Basics
  ☐ SUM, AVERAGE, COUNT
  ☐ CALCULATE, FILTER
  ☐ Time intelligence basics

Week 5: Reports and Design
  ☐ Multi-page reports
  ☐ Slicers, drill-through, tooltips
  ☐ Clean, business-friendly layouts

Week 6: Portfolio Project
  ☐ Build a sales or HR dashboard
  ☐ Publish to Power BI Service
  ☐ Share a public link (or screenshot) in portfolio

Pandas Path — 6 weeks (Data Analyst focus):

Week 1-2: Python Basics
  ☐ Variables, functions, lists, dicts
  ☐ Loops and conditionals
  ☐ Imports and virtual environments

Week 3: Pandas Basics
  ☐ Series and DataFrames
  ☐ read_csv / to_csv
  ☐ Indexing and selecting (loc, iloc)

Week 4: Data Cleaning
  ☐ Handle missing values
  ☐ Fix data types
  ☐ Text cleaning (.str methods)
  ☐ Remove duplicates

Week 5: Transformation
  ☐ groupby and aggregation
  ☐ merge and join
  ☐ pivot tables
  ☐ date and time operations

Week 6: Analysis Project
  ☐ Load a real dataset (Kaggle)
  ☐ Clean it end-to-end
  ☐ Answer 5 business questions
  ☐ Visualise with Matplotlib
  ☐ Write findings in a Jupyter notebook

Next steps after Pandas:
  → SQL (very important for data roles)
  → Matplotlib / Seaborn
  → Power BI or Tableau (visualisation)
  → scikit-learn (if going ML route)