← Home
Calculus Explorer
Δx → dx — Derivative Explorer
geometric · visual · interactive
SPEED
🐢 SLOW FAST 🐇
3
controls all ▶ animations
all pages
SECANT → TANGENT
f(x)
secant Δy/Δx
tangent dy/dx
ghost trails
Δy/Δx=—
dy/dx=—
error=—
angle θ=—
◀ Δx gap (shrinks) remainder (grows) ▶
x₀
0.00
Δx (right pt)
3.00
Δx (left pt)
3.00
DERIVATIVE TRACE f′(x)
5 CURVES + TANGENTS AT x
x
1.00
Δx
0.50
VALUES AT x
curvef(x)f′(x)Δy/Δx
DERIVATIVE FAMILY f′(x)
x value
1.00
f(x)=x² f′(x)=2x
x² DIFFERENCE QUOTIENT
h[f(x+h)−f(x)]/hf′(x)error
f(x)=x³ f′(x)=3x²
x³ DIFFERENCE QUOTIENT
h[f(x+h)−f(x)]/hf′(x)error
f(x)=sin x f′(x)=cos x
sin x DIFFERENCE QUOTIENT
h[f(x+h)−f(x)]/hf′(x)error
x₀
0.50
Δx
3.00
FULL VIEW
TRIANGLE ZOOM
Δx
Δy
Δy/Δx
f′(x)
Δx FINITE100%dx LIMIT
Q SLIDES → P (animated dot trail)
P (fixed)
Q (moving)
trail
angle arc
x₀
0.50
SLOPE TABLE AS Q→P
# h sec slope → tan
ANGLE θ OF SECANT (degrees)
NEGATIVE Δx — Q approaches from LEFT
+Δx (right)
−Δx (left)
tangent
x₀
0.50
|Δx|
2.50
+Δx vs −Δx COMPARISON
property+Δx−Δx
BOTH SECANTS ROTATE TO SAME TANGENT
x
0.50
h (for f″ approx)
0.50
concavity: —
f(x) — ORIGINAL
f′(x) — FIRST DERIVATIVE
f″(x) — SECOND DERIVATIVE
LIVE VALUES
quantityvaluemeaning
NUMERICAL f″ APPROXIMATION
Central difference formula:
f″(x) ≈ [f(x+h) − 2f(x) + f(x−h)] / h²
happrox f″exact f″error
OSCULATING CIRCLE (curvature)
R = (1+f′²)^(3/2) / |f″| — radius of curvature
NON-DIFFERENTIABLE FUNCTIONS
f(x)
+Δx secant
−Δx secant
left limit ≠ right limit
x₀
0.00
|h|
2.00
LEFT vs RIGHT LIMIT
hright Δy/Δxleft Δy/Δxsame?
WHY NOT DIFFERENTIABLE HERE
ERROR vs h (LOG SCALE)
x
1.00
LOG ERROR TABLE — ALL 3 FUNCTIONS
hx² errorx³ errorsin error
ERROR BAR CHART (linear)
PROOF
ALGEBRAIC STEPS
STUDENT CHALLENGE — GUESS THE SLOPE
CHALLENGE
your answer
SCORE
0 / 0
FUNCTION
x position
0.80
Δx (for f″ approx)
0.50
CONCAVE UP ∪
f(x) = —
f′(x) = —
f″(x) = —
κ = —
① f(x) — ORIGINAL green=concave up red=concave down
f(x)
concave up (f″>0)
concave down (f″<0)
tangent
② f′(x) — FIRST DERIVATIVE slope of f(x)
f′(x)
f′ increasing → f concave up
f′ decreasing → f concave down
③ f″(x) — SECOND DERIVATIVE sign determines concavity
f″(x) exact
numerical approx (finite h)
inflection pts (f″=0)
④ CURVATURE COMB κ = |f″| / (1 + f′²)^(3/2)
f(x)
curvature spines (length ∝ κ)
concave-up spines
concave-down spines
LIVE VALUES AT x
quantityvalueinterpretation
📐 CONCAVITY THEORY
f″(x) > 0 → curve bends UPWARD ∪ → slope increasing
f″(x) < 0 → curve bends DOWNWARD ∩ → slope decreasing
f″(x) = 0 → possible INFLECTION POINT
κ = |f″| / (1+f′²)^(3/2)
measures how sharply the curve bends
2nd Derivative Test:
f′(c)=0 & f″(c)>0 → local MIN
f′(c)=0 & f″(c)<0 → local MAX
FUNCTION
point a
-1.50
point b
1.50
λ (mix)
0.50
CONVEX ∪
chord ≥ f: ✓
Jensen gap = —
support: ✓
① CHORD TEST chord lies above curve → convex
f(x)
chord [a,b]
f(λa+(1−λ)b) — curve pt
λf(a)+(1−λ)f(b) — chord pt
② JENSEN'S INEQUALITY f(λa+(1−λ)b) ≤ λf(a)+(1−λ)f(b)
f(x)
f(midpoint) — lower
chord value — upper
Jensen gap (vertical)
③ SUPPORTING LINE tangent lies below curve → convex
f(x)
tangent at λ-point (supporting line)
f(x) ≥ tangent (convex region)
④ f″(x) SIGN — GLOBAL CONVEXITY f″≥0 everywhere → globally convex
f″(x) ≥ 0 → convex region
f″(x) < 0 → concave region
inflection (f″=0 sign change)
LIVE VALUES
quantityvaluenote
📐 CONVEXITY THEORY
GEOMETRIC (chord):
f is convex on [a,b] iff for all λ∈[0,1]:
f(λa+(1−λ)b) ≤ λf(a)+(1−λ)f(b)
i.e. chord lies ON or ABOVE the curve
DIFFERENTIAL (2nd deriv):
f convex on I ↔ f″(x) ≥ 0 ∀x∈I
f concave on I ↔ f″(x) ≤ 0 ∀x∈I
strictly convex if f″(x) > 0 everywhere
SUPPORTING LINE:
f convex ↔ tangent at every pt lies
below or on the curve:
f(x) ≥ f(c) + f′(c)(x−c) ∀x
CONCAVITY vs CONVEXITY:
Concave up = convex (same thing)
Concave down = concave (same thing)
Convention differs by field!
🎵 SPECIAL f′(x) VALUE — triggers alarm tone
when f′(x) =
tolerance ±
When you click a point where f′(x) matches this value,
a special alarm tone + pink highlight triggers.
At f′=0 → lowest audible pitch. Higher f′ → higher pitch.
FREQUENCY MAP (continuous)
f′(x) → frequency (continuous, not fixed)
f′ = 0 → 110 Hz (A₂, audible bass)
f′ = + → rises up to ~1400 Hz
f′ = − → falls down to ~60 Hz
Waveform shape morphs with frequency:
low → sine | mid → triangle | high → sawtooth
CLICK ANY POINT → HEAR ITS SOUND
Click anywhere on the curve
LAST CLICKED POINT
— click the graph —
LIVE WAVEFORM
PITCH RULE
f′(x)=0 → 110 Hz (base audible). Pitch rises/falls continuously with f′.
Waveform morphs: low Hz=sine, mid=triangle, high=sawtooth
2x+1: f′=2 always → same pitch everywhere
Tone sustains until you press ■ STOP
HISTORY — last 6 clicks
🔴 SPECIAL f(x) VALUE
f(x) =
±
🔔 RED FLASH
🔵 SPECIAL f′(x) VALUE
f′(x) =
±
🔔 CYAN FLASH + BEEP
🟢 SPECIAL f″(x) VALUE
f″(x) =
±
🔔 GREEN FLASH
⑬ ALL SOUNDS — f, f′, f″ SONIFICATION — choose a function and click a type —
x₀ (global)
1.00
wave shape
MASTER WAVEFORM
Chord: —
PITCH MAPPING — y=0 always = 220 Hz · frequency rises/falls with y value
y = 0 → 220 Hz (always audible base)
y > 0 → pitch rises above 220 Hz
y < 0 → pitch falls below 220 Hz
Waveform morphs: low=sine · mid=triangle · high=sawtooth · very high=square
f′=2x → rises with x · f″=2 (const 440Hz+)
f′=3x² → always ≥0, low at x=0
sin x f′=cos x → oscillates ±1 around 220Hz
2x+1 f′=2 always → constant note (flat slope!)
Sweep range: x = −20 → +20
Each function sweeps its full behavior — steep functions show dramatic pitch changes.
Set a special f′ value above to hear an alarm when the sweep hits that slope.
SWEEP PLAYER — hear f, f′, f″ sweep across x from left to right
sweep type
sweep speed
🔴 SPECIAL f(x) VALUE
f(x) =
±
→ 🔴 RED FLASH
🔵 SPECIAL f′(x) VALUE
f′(x) =
±
→ 🔵 CYAN FLASH
🟢 SPECIAL f″(x) VALUE
f″(x) =
±
→ ✦ GOLDEN FLASH
☀ Brightness
OFF
☀♪ Brightness + Audio
OFF
function
driven by
color
GRAPH — drag cursor to scrub x
x
0.00
LIVE VALUES
f(x) =
f′(x) =
f″(x) =
BRIGHTNESS
0%
mapped from normalised value of selected derivative
How it works:
Toggle ON → screen flashes to show the chosen derivative's value at x.
Bright = large positive.
Dark = near zero.
Medium = negative (inverted).

Drag x-slider to scrub. Try f″ on sin x — watch brightness pulse!
AUTO SWEEP — brightness animates as x sweeps left → right
speed
sweep type
⑰ CONVERGENCE RACE — same Δx applied to both curves simultaneously
A x₀ 1.00
vs
B x₀ 1.00
Δx (shared) 3.000
Quick presets:
💡 The Journey to dx: For linear functions (2x+1), the secant line is always the tangent — zero error at any Δx. The journey is instant. For x·sin(1/x), the curve oscillates so wildly near x₀ that the secant keeps changing direction — it needs an extremely tiny Δx before it settles. This is the maximum possible gap between Δx and dx.
A — SECANT → TANGENT
Δx=2.0 (huge)
Δx=0.5
Δx=0.05 ≈dx ✓
Δx FINITE ≈ dx
Δy/Δx=—
dy/dx=—
|error|=—
B — SECANT → TANGENT
Δx=2.0 (huge)
Δx=0.5
Δx=0.05 ≈dx ✓
Δx FINITE ≈ dx
Δy/Δx=—
dy/dx=—
|error|=—
A — JOURNEY: HOW MANY STEPS TO REACH dx?
B — JOURNEY: HOW MANY STEPS TO REACH dx?
ERROR RACE — |Δy/Δx − dy/dx| vs Δx  (left = Δx→0 = dx limit)
A — error
B — error
current Δx
≈ dx threshold
LIVE COMPARISON
Δx A |err| B |err| ≈dx?
VERDICT
— move Δx or press ▶ SHRINK BOTH —
⑱ LAGRANGE'S MEAN VALUE THEOREM
↕ drag the a / b / c points directly on the curve (c is constrained to stay between a and b) · tap any point for its exact coordinates
f(x)
secant (a → b)
tangent at c
slope=(f(b)−f(a))/(b−a)=—
f′(c)=—
c=—
A = (—, —)
B = (—, —)
C = (—, —)
a
-1.50
b
1.50
c
0.00
DERIVATIVE f′(x) — horizontal line = secant slope
Integration Sandbox
Drag · Switch · Discover — every number live
🐢 Anim speed
Strip Builder — slice the area into rectangles
Strips n 8
Function
Rule
Approx area
Exact area
Error %
Strip Δx
Σ f(xᵢ)·Δx n= Δx= sum=
Signed Area — teal above, rose below x-axis
Function
Upper limit b 3.14
∫ signed
+ area
− area
F(x) = ∫₀ˣ f — area becoming a curve
Function
x = 1.5
f(x) height
F(x) = area so far
🚗 Velocity → Distance — shaded blocks = metres
v(t) — speed m/s  ·  area = distance
s(t) = ∫v dt — distance metres
Speed profile
Time t = 2.5 s
Speed now
Distance = ∫v dt
Avg speed
Blocks shown
Limits a & b — drag both, watch area change
a = 0.0
b = 2.0
∫ₐᵇ f dx
b − a
Mean value f̄
Error vs n — all methods race toward zero
Fixed demo: f(x) = x² on [0, 3] (exact area = 9) — shows convergence rate comparison
Error @ n=10
Error @ n=50
Error @ n=200
∂∫ The Connection
Derivative & Integration are inverse operations — watch it live
Current Stage
Stage 1 — Accumulate
Slide x to watch the integral accumulate area under f(t) from 0 to x. The height of the accumulation curve at each x equals f(x).
f(t) — the original function
Function
Upper limit x = 1.57
Area A(x)
f(x) (height)
dA/dx ≈ f(x)?
A(x) — area accumulation & its derivative
A(x) = ∫f(t)dt  ·  A′(x) = d/dx A(x)  ·  f(x) original
A(x)
A′(x)
f(x)
A′(x) − f(x)
FTC Part 2 — Antiderivative Undoes Integration
d/dx [ ∫₀ˣ f(t) dt ] = f(x)  ·  ∫ₐᵇ F′(t) dt = F(b) − F(a)
∫₀ˣ f(t)dt
F(x) antideriv
∫₀ˣ f(t)dt
Match?
🦴 WHY INTEGRATION MUST SUBTRACT — CAVEMAN UNDERSTAND
Curve:
a
-1.0
b
2.0
F(b) alone
F(a) junk
F(b)−F(a) ✓
∫ₐᵇ f dt
F(b)−F(a) = ∫ₐᵇ f dt ✓
error: —
🦣 UGH. CAVE-MAN EXPLAIN WHY SUBTRACT.
1
CAVE-MAN FIND ANTIDERIVATIVE F(x).
F(x) like bucket. It hold accumulated stuff since beginning of time (x=−∞). F(x) not know where cave-man want to start counting. F always carry old junk from before.
2
∫ₐᵇ f(t) dt = AREA BETWEEN a AND b ONLY.
Cave-man say: "UGH! Me not care about stuff before a. Me only care about area from a to b." But F(b) has junk from before a inside it. F(a) is exactly the same junk. Same junk! So cave-man SUBTRACT F(a) to cancel junk.
3
F(b) − F(a) = JUNK CANCEL. ONLY GOOD AREA REMAIN.
F(b) = stuff up to b  ·  F(a) = same old junk up to a  ·  F(b)−F(a) = only the area between a and b. Pure! No junk! This why we always subtract.
🪨 THE ROCK-SOLID RULE:
∫ₐᵇ f(t) dt  =  F(b) − F(a)
It not matter which antiderivative F you pick.
All antiderivatives differ only by constant C.
  → [F(b)+C] − [F(a)+C]  =  F(b) − F(a)  ←  C vanish! POOF!
Cave-man very happy. Constant C always die in subtraction. 🦴
SIMPLE CAVE-BRAIN ANALOGY:  You start at position F(a) on number line. You walk to position F(b). Distance walked = F(b) − F(a). F(b) alone tell cave-man WHERE he end up, not HOW FAR he walk. Subtraction give distance. Area = distance F travel. ∫ₐᵇ f dt = F walks from a to b = F(b) − F(a). ALWAYS SUBTRACT. NO EXCEPTION. CAVE-MAN LAW.
Stage 1 — Accumulating area: Drag the x slider left/right. The gold shaded region is the definite integral ∫₀ˣ f(t)dt. Notice: when f(x) is positive, A(x) increases; when negative, it decreases. The slope of A(x) at any x equals f(x) — that's the Fundamental Theorem.
Graph
ESC to close · interactions work in fullscreen
Ad space