Data Visualisation Using Python Training Course

Participants can clean, transform and validate data before creating visualisations.

• Column names standardised using clear, consistent and machine-readable conventions
• Missing values classified as absent, unknown, inapplicable or data-entry errors
• Dates parsed, validated and converted into reporting periods and intervals
• Categorical values normalised through mapping tables and controlled labels
• Outliers investigated using business rules, distribution checks and source records
• Tidy data principles applied before aggregation, charting and automated reporting

Participants can select chart forms and visual encodings that fit the analytical task.

• Nominal, ordinal, interval and ratio data matched to visual encodings
• Bar charts used for ranked categorical comparisons and part-to-whole analysis
• Line charts used for continuous time series and changing performance
• Scatter plots used to examine relationships, clusters and unusual observations
• Position, length, area and colour evaluated through perceptual accuracy
• Chart junk removed while retaining labels, context and decision-relevant detail

Participants can build controlled Matplotlib figures with readable axes, labels and layouts.

• Figure and Axes objects created using Matplotlib object-oriented syntax
• Axis limits selected to avoid distortion and support valid comparisons
• Titles, subtitles and source notes written for specific stakeholder questions
• Tick labels formatted for dates, currencies, percentages and large numbers
• Gridlines, spines and whitespace adjusted to improve chart readability
• PNG, SVG and PDF exports configured for reports and presentations

Participants can visualise category comparisons and composition without obscuring differences.

• Sorted horizontal bar charts built for long category labels
• Grouped bars compared carefully when categories and series remain limited
• Stacked bars used for composition while preserving total comparison context
• Small multiples used to compare repeated measures across business units
• Reference lines added for targets, averages, thresholds and benchmarks
• Annotations used to identify material movements and decision-relevant exceptions

Participants can communicate distributions, variation and group differences using appropriate statistical charts.

• Histograms configured with meaningful bin widths and comparable scales
• Kernel density estimates interpreted cautiously alongside underlying sample sizes
• Box plots used to show median, spread, skewness and outliers
• Violin plots assessed for distribution shape and potential interpretation risks
• Strip plots and swarm plots reveal individual observations within groups
• Confidence intervals displayed when summarising sample-based estimates and comparisons

Participants can use Seaborn to create statistically informed charts from tidy datasets.

• Seaborn themes configured for consistent typography, grids and colour treatment
• catplot used to generate categorical charts through figure-level interfaces
• relplot used for faceted relationship charts and grouped observations
• displot used for histograms, densities and empirical cumulative distributions
• regplot and lmplot used to display fitted relationships responsibly
• Facet grids compare segments while preserving shared scales and labels

Participants can visualise relationships while distinguishing correlation from causal claims.

• Scatter plots encode additional variables through size, colour and facets
• Overplotting reduced using transparency, jitter, binning and hexagonal aggregation
• Correlation heatmaps formatted with ordered variables and readable annotations
• Regression lines accompanied by uncertainty bands and appropriate caveats
• Residual plots used to inspect model fit and systematic error
• Confounding risks identified before presenting relationship charts as evidence

Participants can build time-series charts that show trends, seasonality and operational change.

• Time indexes created from validated dates, frequencies and reporting calendars
• Daily data aggregated into weekly, monthly and quarterly reporting periods
• Rolling averages calculated to reduce noise and reveal underlying trends
• Seasonality compared using indexed periods, small multiples and heatmaps
• Events, policy changes and incidents annotated directly on timelines
• Forecasts separated visually from actuals using labels and distinct styling

Participants can create geographic visualisations while applying appropriate geographic and ethical controls.

• Geographic identifiers checked against reference codes and boundary definitions
• Choropleth maps matched to rates, ratios and standardised measures
• Raw counts avoided when geographic areas have materially different populations
• Colour scales selected to distinguish sequential, diverging and categorical data
• Map classifications tested for sensitivity to breaks and outliers
• Location privacy risks assessed before publishing granular geographic information

Participants can apply visual design standards that improve readability and accessibility.

• Sequential, diverging and qualitative palettes selected for data meaning
• Colourblind-safe palettes tested using contrast and redundant visual cues
• Typography hierarchy applied through title, subtitle, annotation and source-note levels
• Direct labels preferred when legends create unnecessary eye movement
• Minimum contrast ratios considered for screen-based chart consumption
• Accessible descriptions written for charts distributed through digital channels

Participants can structure visual outputs around findings, evidence and required decisions.

• Message titles state the finding rather than merely naming measures
• Annotations explain turning points, exceptions, targets and material differences
• Contextual benchmarks distinguish expected variation from actionable underperformance
• Narrative sequence moves from question through evidence to recommended action
• Chart collections organised for executive, operational and technical audiences
• Limitations documented where data coverage, quality or methods constrain interpretation

Participants can build interactive Plotly charts that support controlled stakeholder exploration.

• Plotly Express creates interactive charts from tidy DataFrames
• Hover templates display precise values, units and relevant contextual fields
• Dropdowns and buttons switch measures without changing analytical definitions
• Range sliders support time-series inspection across extended reporting periods
• Facets and animation used selectively to avoid interaction overload
• Interactive chart exports shared through HTML files and approved platforms

Participants can customise Plotly figures for multi-layered and stakeholder-specific reporting needs.

• Graph Objects traces combined into multi-layer analytical figures
• Secondary axes used only when scales and units remain interpretable
• Custom tooltips expose drill-down fields without crowding the display
• Shapes and annotations mark targets, limits and operational events
• Subplots organise related views with shared axes and aligned comparisons
• Figure layouts standardised through reusable templates and configuration dictionaries

Participants can design focused dashboard views that support monitoring and follow-up action.

• Dashboard questions prioritised before selecting metrics, charts and interaction controls
• Key performance indicators paired with trends, comparisons and explanatory context
• Visual hierarchy directs attention toward material decisions and exceptions
• Filters constrained to prevent conflicting definitions and misleading comparisons
• Dashboard wireframes specify layout, user tasks, sources and refresh frequency
• Reporting governance assigns chart ownership, approval steps and publication controls

Participants can create maintainable Python visualisation code for recurring reporting requirements.

• Plotting functions separate data transformation from visual formatting logic
• Function parameters control titles, measures, filters and output destinations
• Configuration dictionaries store colours, labels, thresholds and chart settings
• Reusable themes enforce consistent fonts, spacing and annotation conventions
• Code comments explain assumptions, transformations and non-obvious design choices
• Version-controlled projects maintain traceable changes to visual reporting code

Participants can review charts for data accuracy, interpretability and publication readiness.

• Source totals reconciled against charted values, filters and aggregations
• Axis scales checked for truncation, inversion and inconsistent intervals
• Labels, units, periods and denominators verified before stakeholder circulation
• Colour meanings tested for consistency across charts and reporting products
• Peer review checklists assess clarity, evidence, accessibility and analytical integrity
• Publication approvals record data owner, reviewer, date and intended audience

Participants can present uncertainty, variability and risk signals without false precision.

• Error bars represent standard errors, confidence intervals or measurement ranges
• Fan charts communicate forecast uncertainty across projected time periods
• Scenario bands compare base, upside and downside planning assumptions
• Control charts distinguish common-cause variation from special-cause signals
• Risk matrices evaluated for ordinal scales, aggregation limits and interpretation issues
• Precision matched to data quality, sample size and decision consequence

Participants can defend visual choices and revise charts in response to stakeholder feedback.

• Stakeholder requests translated into analytical questions, measures and acceptance criteria
• Audience assumptions tested through structured chart walkthroughs and feedback
• Alternative chart designs compared against speed, accuracy and decision relevance
• Technical findings summarised without removing material caveats or uncertainty
• Revision logs record requested changes, rationale and approved final decisions
• Briefing notes identify actions, owners, dates and monitoring indicators

Participants can build, review and present a complete Python visual analysis addressing a defined stakeholder decision.

• Capstone brief defines stakeholder decision, dataset, constraints and success criteria
• Data preparation documents source checks, transformations and analytical assumptions
• Chart suite combines exploratory findings with decision-focused explanatory visuals
• Matplotlib, Seaborn or Plotly selected according to communication requirements
• Quality review tests calculations, accessibility, labels and interpretation risks
• Final presentation explains findings, limitations, recommendations and follow-up actions