Multiple plots (hold on) in MATLAB - Deep Dive

Overview - Multiple plots (hold on)

What is it?

Multiple plots (hold on) in MATLAB allow you to draw several graphs on the same figure without erasing the previous ones. Normally, when you create a new plot, MATLAB clears the old one. Using 'hold on' stops this clearing, so you can add more lines or points to the same graph. This helps compare data visually in one place.

Why it matters

Without 'hold on', you would have to create separate figures for each plot, making it hard to compare data side by side. This wastes time and screen space. 'Hold on' lets you layer multiple data sets on one graph, making patterns and differences easier to see. It is essential for clear, efficient data visualization.

Where it fits

Before learning 'hold on', you should know how to create basic plots in MATLAB. After mastering it, you can learn advanced plotting techniques like subplots, legends, and customizing plot styles to make your graphs more informative and attractive.

Mental Model

Core Idea

Using 'hold on' tells MATLAB to keep the current plot and add new plots on top instead of replacing it.

Think of it like...

Imagine drawing on a transparent sheet over a picture. Without 'hold on', every new drawing erases the old one. With 'hold on', you keep adding new drawings on the same sheet, layering information.

┌───────────────┐
│ Start plot    │
│ (plot command)│
└──────┬────────┘
       │
       ▼
┌───────────────┐
│ hold on       │
│ (keep plots)  │
└──────┬────────┘
       │
       ▼
┌───────────────┐
│ Add new plot  │
│ (plot command)│
└───────────────┘

Build-Up - 7 Steps

1

Foundation

Basic single plot creation

Concept: How to create a simple plot in MATLAB using the plot() function.

Use plot(x, y) where x and y are vectors of the same length. This draws a line graph of y versus x. For example, x = 1:5; y = [2 4 6 8 10]; plot(x, y); creates a line connecting points (1,2), (2,4), etc.

Result

A single line graph appears showing the data points connected.

Understanding how to create a basic plot is essential before adding multiple plots on the same figure.

2

Foundation

Default plot replacement behavior

Concept: By default, each new plot command clears the previous plot in the figure window.

If you run plot(x1, y1) and then plot(x2, y2) without 'hold on', the second plot replaces the first one. This means only the last plot is visible.

Result

Only the second plot is shown; the first disappears.

Knowing that MATLAB clears old plots by default explains why 'hold on' is needed to keep multiple plots visible.

3

Intermediate

Using 'hold on' to combine plots

🤔Before reading on: do you think 'hold on' allows adding plots without clearing the figure, or does it just pause plotting temporarily? Commit to your answer.

Concept: 'hold on' freezes the current figure so new plots add to it instead of replacing it.

After plotting the first data set, type hold on; then plot additional data sets. All plots appear together. Example: x = 1:5; y1 = x; y2 = x.^2; plot(x, y1); hold on; plot(x, y2);

Result

Both lines y = x and y = x^2 appear on the same graph.

Understanding that 'hold on' changes the plot clearing behavior is key to layering multiple data sets visually.

4

Intermediate

Using 'hold off' to reset behavior

🤔Before reading on: does 'hold off' clear the figure immediately or just restore default plotting behavior? Commit to your answer.

Concept: 'hold off' restores the default behavior where new plots replace old ones.

After using 'hold on', calling hold off; means the next plot command will clear the figure first. Example: plot(x, y1); hold on; plot(x, y2); hold off; plot(x, y1); % This clears previous plots

Result

Only the last plot (y = x) is visible after hold off and plotting again.

Knowing how to toggle 'hold on' and 'hold off' helps control when to layer plots and when to start fresh.

5

Intermediate

Adding legends and colors for clarity

Concept: When plotting multiple lines, adding legends and colors helps distinguish them.

Use plot(x, y1, 'r') for red line, plot(x, y2, 'b') for blue line. Add legend('Linear', 'Quadratic') to label lines. Example: plot(x, y1, 'r'); hold on; plot(x, y2, 'b'); legend('Linear', 'Quadratic');

Result

Two colored lines appear with a legend explaining each.

Visual clarity is important when combining plots; colors and legends prevent confusion.

6

Advanced

Plotting different types on one figure

🤔Before reading on: can 'hold on' combine different plot types like scatter and line plots on the same figure? Commit to your answer.

Concept: 'hold on' allows mixing plot types such as lines, scatter points, and bar charts on one figure.

Example: x = 1:5; y1 = x; y2 = x.^2; plot(x, y1, 'r'); hold on; scatter(x, y2, 'filled'); legend('Line', 'Scatter');

Result

A red line and scatter points appear together on the same graph.

Knowing 'hold on' supports multiple plot types expands your visualization options.

7

Expert

Understanding hold state and figure handles

🤔Before reading on: does 'hold on' affect all figures globally or only the current figure? Commit to your answer.

Concept: 'hold on' applies only to the current figure and axes, controlled by figure and axes handles internally.

MATLAB tracks hold state per figure. Using figure handles, you can control which figure holds plots. Example: fig1 = figure; plot(x, y1); hold on; plot(x, y2); fig2 = figure; plot(x, y2); % This figure is independent hold on; plot(x, y1);

Result

Two separate figures each hold their own plots without interference.

Understanding hold state per figure prevents confusion when working with multiple figures and complex plotting.

Under the Hood

'hold on' changes an internal flag in the current figure's axes that tells MATLAB not to clear the axes before plotting new data. Normally, each plot command resets the axes, but with hold on, MATLAB appends new graphics objects to the existing axes. This flag is stored in the figure's graphics object properties and affects rendering order and layering.

Why designed this way?

MATLAB was designed to clear plots by default to avoid clutter and confusion for simple use cases. 'hold on' was added as an explicit command to give users control when layering plots is needed. This design balances simplicity for beginners and flexibility for advanced users.

Current Figure
┌─────────────────────────────┐
│ Axes Object                 │
│ ┌───────────────────────┐ │
│ │ hold_state flag        │ │
│ │ (on/off)              │ │
│ └─────────┬─────────────┘ │
│           │               │
│           ▼               │
│  Plot command checks hold │
│  state:                   │
│  ┌─────────────────────┐  │
│  │ If hold on: append   │  │
│  │ new plot to axes     │  │
│  │ Else: clear axes     │  │
│  │ and draw new plot    │  │
│  └─────────────────────┘  │
└─────────────────────────────┘

Myth Busters - 4 Common Misconceptions

Quick: Does 'hold on' keep plots forever or only until 'hold off' is called? Commit to yes or no.

Common Belief: Once you use 'hold on', all future plots will always add to the figure forever.

Reality: 'hold on' keeps plots only until you call 'hold off' or close the figure. After 'hold off', new plots replace old ones again.

Why it matters: Assuming 'hold on' is permanent can cause unexpected plot clearing and confusion when adding new data.

Quick: Does 'hold on' affect all open figures or just the current one? Commit to one answer.

Common Belief: 'hold on' applies globally to all figures in MATLAB.

Reality: 'hold on' affects only the current active figure and its axes, not other open figures.

Why it matters: Misunderstanding this can lead to bugs when working with multiple figures, expecting plots to layer everywhere.

Quick: Can 'hold on' be used to combine plots with different axis scales automatically? Commit to yes or no.

Common Belief: 'hold on' automatically adjusts axes to fit all plots with different scales.

Reality: 'hold on' does not change axis limits automatically; you must manually adjust axes or use commands like axis tight or ylim to fit all data.

Why it matters: Assuming automatic scaling can cause plots to be clipped or misleading if axis limits are not managed.

Quick: Does 'hold on' affect the order in which plots are drawn? Commit to yes or no.

Common Belief: 'hold on' draws new plots behind existing ones by default.

Reality: New plots are drawn on top of existing plots, which can hide earlier plots if colors or styles overlap.

Why it matters: Knowing draw order helps avoid hiding important data and guides use of transparency or layering.

Expert Zone

1

'hold on' affects only the current axes, so if you create multiple axes in one figure, you must set hold state for each separately.

2

Using 'hold on' can increase memory usage because all plot objects remain in the figure until cleared or closed.

3

Combining 'hold on' with plotting functions that reset axes properties (like plotyy or tiledlayout) requires careful management to avoid unexpected behavior.

When NOT to use

'hold on' is not suitable when you want to create multiple separate plots side by side; use subplot or tiledlayout instead. For dynamic or animated plots, updating data properties is better than layering plots with 'hold on'.

Production Patterns

Professionals use 'hold on' to overlay experimental data with model predictions, add error bars or annotations on the same graph, and combine different data types (lines, scatter, bars) for comprehensive visual analysis.

Connections

Layered image editing

'hold on' is like adding layers in image editing software where each layer adds new content without erasing others.

Understanding layering in image editing helps grasp how multiple plots coexist visually in one figure.

State management in programming

'hold on' changes the internal state of the figure, similar to how state variables control behavior in software applications.

Recognizing 'hold on' as state management clarifies why its effect persists until explicitly changed.

Transparency in painting

Just like painters add transparent layers to build a complex image, 'hold on' lets you add plot layers to build complex visualizations.

This connection shows how layering builds complexity gradually, a principle across art and data visualization.

Common Pitfalls

#1 Forgetting to use 'hold on' when adding multiple plots causes previous plots to disappear.

Wrong approach:

plot(x1, y1);
plot(x2, y2); % This replaces the first plot

Correct approach:

plot(x1, y1);
hold on;
plot(x2, y2); % Both plots appear together

Root cause:

Not knowing that MATLAB clears the figure by default before each new plot.

#2 Calling 'hold on' but never calling 'hold off' leads to unexpected layering in later plots.

Wrong approach:

plot(x1, y1);
hold on;
plot(x2, y2);
plot(x3, y3); % All plots layer forever

Correct approach:

plot(x1, y1);
hold on;
plot(x2, y2);
hold off;
plot(x3, y3); % Last plot replaces previous

Root cause:

Not resetting hold state causes persistent layering beyond intended scope.

#3 Assuming 'hold on' automatically adjusts axes to fit all data causes plots to be clipped.

Wrong approach:

plot(x1, y1);
hold on;
plot(x2, y2); % No axis adjustment

Correct approach:

plot(x1, y1);
hold on;
plot(x2, y2);
axis tight; % Adjust axes to fit all data

Root cause:

Misunderstanding that axis limits must be managed separately from hold state.

Key Takeaways

'hold on' lets you add multiple plots to the same figure without erasing previous ones, enabling layered visualizations.

By default, MATLAB clears the figure before each plot, so 'hold on' changes this behavior to keep plots visible together.

The hold state applies only to the current figure and axes, so managing multiple figures requires careful control.

Using 'hold off' restores default plotting behavior, allowing you to start fresh when needed.

Adding legends, colors, and managing axis limits are important to make multiple plots clear and informative.