Understanding the difference between left and right Riemann sum is a foundational step in mastering integral calculus. On top of that, these methods serve as the primary bridge between discrete summation and continuous integration, allowing us to approximate the area under a curve using rectangles. While both techniques partition an interval into subintervals of equal width, the critical distinction lies in which endpoint of each subinterval determines the rectangle's height. This choice directly impacts whether the approximation overestimates or underestimates the true area, a concept vital for understanding error bounds and the definition of the definite integral And that's really what it comes down to..
The Core Concept: Partitioning the Interval
Before diving into the specific differences, You really need to establish the shared framework. Suppose we want to approximate the area under a continuous function $f(x)$ on a closed interval $[a, b]$. We divide this interval into $n$ subintervals of equal width, denoted as $\Delta x$ Easy to understand, harder to ignore..
$ \Delta x = \frac{b - a}{n} $
The partition creates endpoints $x_0, x_1, x_2, \dots, x_n$ where $x_0 = a$ and $x_n = b$. The general formula for any Riemann sum is:
$ \sum_{i=1}^{n} f(x_i^*) \Delta x $
The variable $x_i^$ represents the sample point chosen within the $i$-th subinterval $[x_{i-1}, x_i]$. **The difference between left and right Riemann sum boils down entirely to how $x_i^$ is selected.**
Left Riemann Sum: The Leading Edge
In a Left Riemann Sum (often denoted as $L_n$ or $LHS$), the height of each rectangle is determined by the function value at the left endpoint of the subinterval.
Formula: $ L_n = \sum_{i=1}^{n} f(x_{i-1}) \Delta x $ Or explicitly: $ L_n = [f(x_0) + f(x_1) + \dots + f(x_{n-1})] \Delta x $
Visual Interpretation: Imagine drawing rectangles under the curve. For the first subinterval $[x_0, x_1]$, the rectangle stretches from $x_0$ to $x_1$ with a height of $f(x_0)$. The top-left corner of the rectangle touches the curve Easy to understand, harder to ignore..
Behavior on Monotonic Functions:
- Increasing Functions: Since $f(x_{i-1}) < f(x)$ for all $x$ in $(x_{i-1}, x_i)$, the rectangles sit below the curve. The Left Riemann Sum underestimates the true area.
- Decreasing Functions: Since $f(x_{i-1}) > f(x)$ for all $x$ in $(x_{i-1}, x_i)$, the rectangles extend above the curve. The Left Riemann Sum overestimates the true area.
Right Riemann Sum: The Trailing Edge
In a Right Riemann Sum (denoted as $R_n$ or $RHS$), the height of each rectangle is determined by the function value at the right endpoint of the subinterval And that's really what it comes down to..
Formula: $ R_n = \sum_{i=1}^{n} f(x_i) \Delta x $ Or explicitly: $ R_n = [f(x_1) + f(x_2) + \dots + f(x_n)] \Delta x $
Visual Interpretation: For the first subinterval $[x_0, x_1]$, the rectangle spans the same width but has a height of $f(x_1)$. The top-right corner of the rectangle touches the curve.
Behavior on Monotonic Functions:
- Increasing Functions: Since $f(x_i) > f(x)$ for all $x$ in $(x_{i-1}, x_i)$, the rectangles extend above the curve. The Right Riemann Sum overestimates the true area.
- Decreasing Functions: Since $f(x_i) < f(x)$ for all $x$ in $(x_{i-1}, x_i)$, the rectangles sit below the curve. The Right Riemann Sum underestimates the true area.
Side-by-Side Comparison
To solidify the distinction, the following table summarizes the difference between left and right Riemann sum across key dimensions:
| Feature | Left Riemann Sum ($L_n$) | Right Riemann Sum ($R_n$) |
|---|---|---|
| Sample Point ($x_i^*$) | Left Endpoint ($x_{i-1}$) | Right Endpoint ($x_i$) |
| Index Range | $i = 0$ to $n-1$ | $i = 1$ to $n$ |
| First Rectangle Height | $f(a)$ | $f(a + \Delta x)$ |
| Last Rectangle Height | $f(b - \Delta x)$ | $f(b)$ |
| Increasing Function | Underestimate | Overestimate |
| Decreasing Function | Overestimate | Underestimate |
| Concave Up | Underestimate (usually) | Overestimate (usually) |
| Concave Down | Overestimate (usually) | Underestimate (usually) |
You'll probably want to bookmark this section.
A Concrete Numerical Example
Let’s approximate $\int_0^2 x^2 , dx$ using $n = 4$ subintervals.
-
Setup:
- $a=0, b=2, n=4$
- $\Delta x = \frac{2-0}{4} = 0.5$
- Endpoints: $x_0=0, x_1=0.5, x_2=1.0, x_3=1.5, x_4=2.0$
- Function values: $f(0)=0, f(0.5)=0.25, f(1)=1, f(1.5)=2.25, f(2)=4$
-
Left Riemann Sum ($L_4$): Uses $x_0, x_1, x_2, x_3$. $ L_4 = [f(0) + f(0.5) + f(1) + f(1.5)] \times 0.5 $ $ L_4 = [0 + 0.25 + 1 + 2.25] \times 0.5 = 3.5 \times 0.5 = \mathbf{1.75} $ True value is $\frac{8}{3} \approx 2.67$. Since $x^2$ is increasing on $[0,2]$, $L_4$ is an underestimate.
-
Right Riemann Sum ($R_4$): Uses $x_1, x_2, x_3, x_4$. $ R_4 = [f(0.5) + f(1) + f(1.5) + f(2)] \times 0.5 $ $ R_4 = [0.25 + 1 + 2.25 + 4] \times 0.5 = 7.5 \times 0.5 = \mathbf{3.75} $ As predicted for an increasing function, $R_4$ is an overestimate.
Notice that the true area ($2.67$) lies neatly between the two approximations ($1.75 < 2.That said, 67 < 3. Consider this: 75$). This bracketing property is a powerful tool for error estimation Which is the point..
The Role of Concavity: Beyond Monotonicity
While monotonicity (increasing/decreasing) dictates the behavior of left vs. right sums perfectly, **
