How to Turn Real‑Estate Data into Insightful Reports and Charts with Pandas

Introduction

Reports and statistical charts are essential for presenting data insights. Using a second‑hand housing CSV file, this article shows step‑by‑step how to generate different types of reports and corresponding visualisations.

Report Types

Reports are classified by the variables they contain:

Dimension (categorical) indicators

Measure (continuous) indicators

Only dimension indicators produce frequency tables (single‑category) or cross‑tables (multiple categories). When both dimensions and measures are present, the result is a summary table , where measures appear as statistical aggregates such as mean, sum, or count.

Bar Charts and Their Relation to Tables

Bar charts map directly to tables: a one‑dimensional bar chart corresponds to a single categorical variable, while a two‑dimensional bar chart represents two categorical variables. The bar length reflects a frequency or a statistical measure.

Dataset Overview

The example uses sndHsPr.csv, which contains fields like district, price, subway proximity, and school‑district status. The goal is to explore both the statistical characteristics of the houses and the factors influencing those characteristics.

1) Single‑Factor Frequency

Analyze the distribution of a single categorical variable, providing count, percentage, and cumulative values.

Code example: snd.district.value_counts() To visualise the frequencies as a bar chart:

snd.district.value_counts().plot(kind='bar')

2) Table Analysis (Cross‑Table)

Examine the joint distribution of two categorical variables, showing cell frequencies, percentages, and marginal distributions.

Generate a cross‑table with pandas: pd.crosstab(snd.subway, snd.school) Standardise the rows to percentages and plot a stacked bar chart:

sub_sch = pd.crosstab(snd.subway, snd.school)
row_perc = sub_sch.div(sub_sch.sum(axis=1), axis=0)
row_perc.plot(kind='bar', stacked=True)

The article recommends a custom stack2dim function for creating stacked bar charts. Its key parameters are: raw: the pandas DataFrame i, j: the names of the two categorical variables (as quoted strings, e.g., "school")

3) Summary Statistics

Group data by a categorical variable and compute descriptive statistics for a continuous variable.

Example for average, maximum, and minimum house prices per district:

snd.price.groupby(snd.district).agg(['mean', 'max', 'min'])

Visual Illustrations

The article includes several figures that illustrate each step, such as the two‑dimensional table template, single‑factor frequency chart, cross‑table visualisation, stacked bar chart, and summary statistics chart.