---
title: Creating Graphs
description: Guide to creating visualizations with Graph Builder. Covers histograms, scatter plots, bar charts, time series plots, pair plots, and datetime histograms with interactive selection and export options.
---

# Creating Graphs {#creating-graphs}

Create graphs to visualize your data. This page explains basic graph types including histograms, scatter plots, bar charts, and time series plots.

## How to Create Graphs {#how-to-create-graphs}

1. **Open Graph Builder tab**: Select **Analysis** → **Graph Builder...** from the menu bar
2. **Select dataset**: Choose the dataset to graph from the dropdown menu in the header
3. **Select graph type**: Click the type of graph you want to display (Histogram, Scatter Plot, Bar Chart, etc.)
4. **Configure settings**: Set axes and variables according to the selected graph type
   - Histogram: Variable, number of bins, grouping
   - Scatter plot: X-axis, Y-axis, color coding
   - Bar chart: Category, value, aggregation method, sorting
5. **Adjust options**: Set filter conditions and additional options as needed

Graphs are displayed in real-time in the preview area on the right, allowing you to explore optimal visualizations while changing settings.

## Basic Graphs {#basic-graphs}

### Histogram {#histogram}

Displays the distribution of a single numeric variable as a bar graph. Useful for understanding frequency distribution and shape of data, and identifying outliers.

**Use cases**:

- Check age distribution
- Examine test score distribution
- Understand sales amount distribution

**Settings**:

Basic tab:

- **Column**: Select the numeric column to graph
- **Bins**: Specify the number of bars (bins)
- **Orientation**: Vertical or horizontal bars
- **Show Density Curve**: Overlay a smooth curve computed by [kernel density estimation](glossary#kde). The density curve is drawn on a second axis (Density) on the right, so the count axis of the bars is unchanged. This shows the shape of the distribution independently of the bin boundaries

Grouping tab:

- **Group By**: Select a category column to display color-coded by group

Annotations tab:

- **Show Mean**: Display the mean as a solid line
- **Show Median**: Display the median as a dashed line

![Histogram of numeric variable distribution created with MIDAS: displays data frequency per bin and visualizes distribution shape and outliers](../shared/images/histogram-example.webp)

### Scatter Plot {#scatter-plot}

Visualizes the relationship between two numeric variables as points. Discover correlations, patterns, clusters, and outliers. Also supports contour display based on 2D [kernel density estimation](glossary#kde) and grouping by categorical variables.

**Use cases**:

- Examine relationship between height and weight
- Check correlation between advertising spend and sales
- View relationship between scores in two exam subjects

**Settings**:

Basic tab:

- **X-Axis**: Numeric column for horizontal axis
- **Y-Axis**: Numeric column for vertical axis
- **X-Axis Scale / Y-Axis Scale**: Axis scale (Linear, Log, or Square Root)
- **Color**: Select a category column for color-coded grouping

Density tab:

- **Display Mode**: Select display mode
  - Points only: Shows points only
  - Density only: Shows only contours from 2D [kernel density estimation](glossary#kde)
  - Points + Density (Outliers Only): Shows contours and points. Despite the label, all points are currently displayed, producing the same result as Points + Density (All Points)
  - Points + Density (All Points): Shows contours and all points

Reference Lines tab:

- **Add Reference Line**: Adds reference lines to mark specific values or thresholds

![MIDAS Scatter Plot: places two numeric variables on X and Y axes, displaying each data point as a dot. Used for discovering correlations and patterns](../shared/images/scatter-plot-example.webp)

### Bar Chart {#bar-chart}

Compares values by category using vertical or horizontal bars. Ideal for comparing categorical data, ranking displays, and showing aggregate statistics (count, sum, average, etc.).

**Use cases**:

- Compare sales by product category
- Display customer count by region
- Aggregate order count by month

**Settings**:

- **Category Column**: Column representing categories
- **Values**: Select a numeric column and an aggregation method (Sum, Average, Min, or Max). Choose Count (Frequency) instead to count rows per category
- **Orientation**: Vertical or horizontal bars
- **Sort Order**: Sort order (alphabetical, value descending/ascending)
- **Top N Categories**: Display only the top N categories
- **Show value labels**: Display values on top of bars

![MIDAS Bar Chart: displays values by category as vertical or horizontal bars, with selectable aggregation methods (count, sum, average, etc.)](../shared/images/bar-chart-example.webp)

### Time Series Plot {#time-series-plot}

Displays data changes over time as a line chart. Identify trends, seasonality, and patterns. Multiple time series are displayed together, distinguished by color.

**Use cases**:

- Check daily sales trends
- Track monthly visitor changes
- Observe annual temperature variations

**Settings**:

Basic tab:

- **X-Axis (Time)**: Column for the horizontal axis. Date and datetime columns can be selected, as well as numeric columns such as elapsed time or sequence numbers
- **Y-Axis (Values)**: Numeric columns to display (multiple selection supported)

Ranges tab:

- **Custom Range Bands**: Add bands to highlight specific periods or ranges

Advanced tab:

- **Chart Title**: Set the graph title

![MIDAS Time Series Plot: displays data changes along time axis as line chart, visualizing trends and seasonal patterns](../shared/images/time-series-example.webp)

## Other Graphs {#other-graphs}

### Pair Plot {#pair-plot}

Displays pairwise relationships between multiple numeric variables as a scatter plot matrix. The diagonal shows histograms, and off-diagonal cells show scatter plots with correlation coefficients.

**Use cases**:

- Check correlations between multiple measurements at once
- Understand the overall structure of a dataset

**Settings**:

- **Variables**: Select numeric columns via checkboxes. Columns with an interval/ratio [scale](data-preparation) are eligible, and 2 to 10 variables can be selected. A warning about increased rendering time is shown when more than 6 variables are selected
- **Display Order**: Change the display order of selected variables

![MIDAS Pair Plot: displays pairwise relationships between multiple numeric variables as scatter plot matrix. Shows distribution of each variable on diagonal, allowing simultaneous confirmation of multivariate correlations](../shared/images/pair-plot-example.webp)

### Datetime Histogram {#datetime-histogram}

Visualizes the distribution of datetime data as a histogram on a time axis. Aggregates event frequency by time interval to identify temporal patterns.

**Use cases**:

- Check log event frequency by time period
- Visualize concentration of access times
- Analyze temporal distribution of orders and transactions

**Settings**:

- **Date/Time Column**: Column representing datetime or date
- **Time Interval**: Aggregation time interval (Auto, 1 minute to 1 year)
- **Show trend line**: Display a trend line using [LOESS (locally weighted regression)](glossary#loess). The span is fixed at 0.75 (75% of all data points are used as the neighborhood)

![MIDAS Datetime Histogram: displays distribution of datetime data as histogram on time axis, visualizing event frequency and temporal patterns](../shared/images/datetime-histogram-example.webp)

### Custom Graph {#custom-graph}

Creates flexible visualizations based on Grammar of Graphics principles. Build layered plots by combining geometric elements, statistical transformations, and aesthetic mappings.

See the [Custom Graph Guide](custom-graph) for details.

## Graph Operations {#graph-operations}

### Interactive Selection {#interactive-selection}

All graphs support data point selection. Selection methods vary by graph type:

**Click selection**: Click data points, bars, or bins to select. Ctrl/Cmd+click to add to the selection.

**Rectangle selection**: Switch to Select mode using the mode toggle buttons in the top-right corner of the graph, then drag to select.

- **Pan mode**: Drag graph to move the display range (default)
- **Select mode**: Drag mouse to select a range

Selected rows are highlighted in other tabs such as Data Table, Statistics, and Selected Rows. See [Row Selection](row-selection) for details on cross-tab synchronization.

### Export {#export}

Click the **Export as SVG** button below the preview to download the graph as an SVG file. SVG is a vector format that maintains quality at any scale.

![Export Graph dialog](../shared/images/export-graph-modal.webp)

**File Name** - Specify the output file name. The `.svg` extension is added automatically.

**Export Size** - Specify the width and height of the output SVG in pixels. Values range from 100 to 10000. The initial values match the current preview size.

**Aspect ratio lock** - Locks the aspect ratio between width and height. Changing one value automatically adjusts the other.

Graphs added to a report can also be exported from the report element menu. See [Reports](report#export) for details.

### Saving Graphs {#saving-graphs}

Graph settings created in Graph Builder are automatically saved to the project as tab state. Graph settings are preserved when you close and reopen the project.

### Changing Settings {#changing-settings}

Graph settings (column selection, filters, display options, etc.) can be changed at any time. Graphs update in real-time when settings are changed.

## Next steps {#next-steps}

- **[Custom Graph](custom-graph)** - Flexible visualization with Grammar of Graphics
- **[Report](report)** - Compile graphs and tables into reports

## See also {#see-also}

- **[Basic Statistics](basic-statistics)** - Scatter plot matrices and correlation matrices for multivariate comparison
- **[Data Preparation and Import](data-preparation)** - Measurement scale settings that affect graph creation