Search This Blog

Data visualizations

Here is a list of 100 command prompts for creating a wide variety of data visualizations, categorized by their analytical purpose.

1. Comparison (Comparing values and categories)

  1. Create a vertical bar chart of [total sales] for each [product category].

  2. Generate a horizontal bar chart to rank [countries] by [population].

  3. Plot a grouped bar chart to compare [sales] and [profit] for each [region].

  4. Generate a diverging bar chart to show [profit margin] for [products], centered at zero.

  5. Create a lollipop plot to compare [metric Y] for [category X] (for a less cluttered look).

  6. Plot a bar-line combo chart to show [sales (bars)] and [growth rate (line)] on the same plot.

  7. Create a bullet graph for [KPI] showing [current value] vs. [target value] and [performance ranges].

  8. Compare [Group A] and [Group B] using side-by-side box plots.

  9. Create a radar chart to compare [N features] for [three different products].

  10. Plot a slopegraph to compare the change in [ranking] for [items] between [Time 1] and [Time 2].

  11. Generate a dot plot to compare [values] across [categories] with minimal clutter.

  12. Create a table-bar chart (bar plot within a table) to show [metric] for each [item].

2. Distribution (Visualizing the spread of data)

  1. Plot a histogram of [age] to see its distribution.

  2. Generate a density plot (KDE) for [price] to see the shape of its distribution.

  3. Create a histogram with [20] bins for [variable].

  4. Overlay a normal distribution curve on the histogram of [data].

  5. Plot a box plot (box-and-whisker) for [salary] to identify median, quartiles, and outliers.

  6. Generate a violin plot for [test scores] by [class] to show both the distribution shape and the box plot.

  7. Create a swarm plot (strip plot) for [data] (best for small datasets).

  8. Plot an Empirical Cumulative Distribution Function (ECDF) for [variable].

  9. Generate a Q-Q plot to check if [data] follows a normal distribution.

  10. Create a rug plot to show the distribution of individual points on a single axis.

  11. Plot a histogram of the [log-transformed variable] to handle heavy skew.

  12. Generate a joyplot (ridgeline plot) to show the distribution of [metric] for [multiple categories].

3. Relationship (Exploring links between variables)

  1. Create a scatter plot of [variable X] (e.g., ad spend) vs. [variable Y] (e.g., sales).

  2. Add a linear regression line (trendline) to the scatter plot of [X] and [Y].

  3. Generate a 2D density plot (heatmap) for a large scatter plot to show point density.

  4. Plot a bubble chart using [X-axis], [Y-axis], and [Z-variable (as bubble size)].

  5. Color the scatter plot points based on [categorical variable] (e.g., 'Region').

  6. Create a scatter plot matrix (pairs plot) for [list of variables] to see all pairwise correlations.

  7. Generate a correlation matrix (heatmap) for all numerical features.

  8. Plot a hexbin plot for [X] and [Y] to handle overplotting in large datasets.

  9. Create a contour plot to show 3D data (X, Y, Z) in 2D.

  10. Plot a 3D scatter plot with [X], [Y], and [Z] axes.

  11. Generate a connected scatter plot to show the relationship between [X] and [Y] over [time].

  12. Plot a co-occurrence matrix (heatmap) for [words/items].

4. Composition (Showing part-to-whole breakdowns)

  1. Create a pie chart showing the [market share] of [companies].

  2. Generate a doughnut chart for [metric] broken down by [category].

  3. Plot a 100% stacked bar chart to show the [percentage composition] of [categories] for each [group].

  4. Create a 100% stacked area chart to show the change in [composition] over [time].

  5. Use a treemap to visualize [file sizes] within [folders] (hierarchical composition).

  6. Generate a sunburst chart to show hierarchical data for [category] and [subcategory].

  7. Create a waffle chart (square pie chart) to show [percentage] as a grid.

  8. Plot a mosaic plot to visualize the relationship between [two categorical variables].

  9. Generate a stacked bar chart (non-100%) to show [total] and [components] for each [category].

5. Change Over Time (Visualizing time-series data)

  1. Plot a line chart of [stock price] over the [last year].

  2. Generate a multi-line chart to compare [metric] for [Group A] and [Group B] over [time].

  3. Create a stacked area chart to show [total sales] broken down by [region] over [time].

  4. Plot a step chart for [data] that changes at discrete intervals (e.g., interest rates).

  5. Generate a candlestick chart for [stock data] (Open, High, Low, Close).

    Shutterstock
  1. Create a Gantt chart to visualize [project tasks] and [timelines].

  2. Plot a streamgraph to show [topic] volume over [time] (a variant of a stacked area chart).

  3. Decompose the [time-series variable] into [trend], [seasonality], and [residual] plots.

  4. Plot an Autocorrelation (ACF) plot for the [time series].

  5. Plot a Partial Autocorrelation (PACF) plot for the [time series].

  6. Create a heatmap of [metric] by [day of week] and [hour of day].

  7. Generate a horizon graph for [N time series] in a compact space.

  8. Plot a calendar heatmap to show [daily metric] (e.g., steps) over a [year].

  9. Create a cycle plot to visualize seasonal patterns over time.

6. Geospatial (Plotting data on maps)

  1. Create a choropleth map of [USA] shaded by [population density].

  2. Generate a bubble map showing [city locations] with [population (as bubble size)].

  3. Plot a connection map (arc map) showing [flight routes] from [origin] to [destination].

  4. Create a simple scatter plot on a map using [latitude] and [longitude] points.

  5. Generate a heatmap layer on a map to show the [density] of [events] (e.g., taxi pickups).

  6. Plot a cartogram based on [metric] (distorting geography to show size).

  7. Create a dot density map for [population distribution].

  8. Plot [N] small multiples maps, one for each [year/category].

7. Flow & Process (Showing movement and funnels)

  1. Generate a Sankey diagram to show the [flow] from [source] to [destination] (e.g., user journey, energy flow).

  2. Create a funnel plot to visualize [drop-off rates] in the [sales/conversion] process.

  3. Plot a dependency graph (network diagram) of [nodes] and [edges].

  4. Generate a chord diagram to show [relationships] and [flow] between [entities].

  5. Create a simple flowchart for the [business process].

  6. Plot a directed network graph (digraph) to show [process flow].

  7. Generate a waterfall chart to show the cumulative effect of [positive/negative] changes.

8. Statistical & Model-Based (For ML and stats)

  1. Plot the ROC curve and calculate the AUC for the [classification model].

  2. Generate a confusion matrix as a heatmap.

  3. Plot the Precision-Recall (PR) curve.

  4. Create a plot of [residuals] vs. [predicted values] for the [regression model]. 7E. Plot the [Elbow Method] curve to find the optimal [k] for [K-Means].

  5. Generate a silhouette plot for the [clustering] results.

  6. Plot the [learning curve] for the [model] showing [train] vs. [validation] score.

  7. Plot the [feature importance] from the [Random Forest model] as a horizontal bar chart.

  8. Generate a dendrogram for the [hierarchical clustering] model.

  9. Plot a 2D decision boundary for the [classifier].

  10. Plot a partial dependence plot (PDP) for [feature] in the [model].

9. Text Visualization (For NLP)

  1. Create a word cloud from [text data] (weighted by frequency).

  2. Generate a bar chart of the top [20] most frequent [words/n-grams].

  3. Plot a t-SNE or UMAP visualization of [document/word] clusters.

  4. Create a heatmap of [keyword] frequency by [document].

  5. Plot a network graph of [character/word] co-occurrences.

  6. Generate a sentiment analysis bar chart showing [positive/negative/neutral] counts.

10. Formatting & Styling (Modifying existing plots)

  1. Set the title of the plot to "[Title]" and add a [subtitle].

  2. Label the X-axis as "[X-Label]" and the Y-axis as "[Y-Label]".

  3. Set the Y-axis to a logarithmic scale.

  4. Add a horizontal (or vertical) line at [value] to mark a [threshold/average].

  5. Annotate the [peak/specific point] on the plot with the text "[Annotation]".

  6. Change the plot style/theme to [e.g., 'seaborn-darkgrid', 'ggplot', 'fivethirtyeight'].

  7. Create a "small multiples" plot (faceted plot) of [charts] broken down by [category].

  8. Rotate the X-axis labels 45 degrees for readability.

  9. Combine [Plot 1] and [Plot 2] into a single dashboard with [2 rows, 1 column].

No comments:

Post a Comment

Agents

  Build me an agent that finds news from WVMR (West Virginia Mountain Radio in Pocahontas County, West Virginia and rewrites as a news st...

Shaker Posts