Summation & Series Explorer

See how each indexed term contributes to a finite sum instead of receiving only a final number. Enter a rational expression in k or define an arithmetic or geometric progression, then choose inclusive index bounds and a positive step. The explorer keeps decimal inputs exact, shows partial sums and positive/negative cancellation, and lets you inspect any prefix alongside the full result. Exact tables and CSV retain very large or tiny fractions that approximate charts cannot distinguish. Explicit geometric mode also evaluates infinite continuation and its signed remainder when the conditions hold; custom expressions are not given unsupported convergence claims.

Exact sums & series

See how every term shapes the sum.

Explore indexed expressions, arithmetic progressions and geometric series with exact fractions, partial sums and transparent cancellation.

Define the indexed series

Switching modes preserves hidden drafts. Only active expression or parameters are evaluated.

Use k, decimals, parentheses and explicit + - * / ^ operators. Examples: 1/(k*(k+1)), k^2, (-1)^k/(k+1). No functions or implicit multiplication.

Integer from -10000 to 10000; included.

Integer at least the lower index. Up to 2000 selected terms.

Positive integer from 1 to 10000. Stop at the last index no greater than the upper bound.

Exact rational arithmetic. Expressions: 512 characters, 256 tokens, 32 nesting levels; integer powers −2000…2000. Calculation limits: 250,000 rational operations and 2,048 digits per reduced numerator or denominator. Undefined terms invalidate the full requested sum.

20 terms evaluated. Finite sum ≈ 0.952380952381; exact values below.

Exact finite sum

≈ 0.952380952381

Exact: 20/21

20 terms; actual last index 20.

Mean of selected terms

≈ 0.047619047619

Exact: 1/21

Absolute term total

≈ 0.952380952381

Exact: 20/21

Sum of absolute term values, before positive/negative cancellation.

Positive contribution

≈ 0.952380952381

Exact: 20/21

Negative contribution

0 (exact)

Exact: 0

Fraction canceled

0 (exact)

Exact: 0

(Absolute total − absolute sum result) / absolute total; zero for all-zero terms. Not an error bound.

Does this continue to an infinite sum?

This mode evaluates a finite sum only. A graph that appears to level off does not establish convergence. Choose an explicitly geometric progression to use its proven infinite-sum formula.

Inspect a partial sum

Use arrow keys to move one term, or Home/End for the first/last term. The table shows the selected block of 20 terms.

Term at k = 1

≈ 0.5

Exact: 1/2

Sum through position 1

≈ 0.5

Exact: 1/2

Prefix mean

≈ 0.5

Exact: 1/2

Remaining contribution to full finite sum

≈ 0.452380952381

Exact: 19/42

Individual terms

Vertical range 0 (exact) to ≈ 0.5. Solid blue line; gray line marks zero. Selected term is circled.

Values at indices 1 through 20. Use the term selector and table for exact values; the line connects discrete samples and does not evaluate intermediate indices.
k = 1k = 20

Chart positions use normalized exact ratios, then truncated display coordinates. Very small differences can coincide visually; the exact table and CSV preserve them.

Running partial sums

Vertical range 0 (exact) to ≈ 0.952380952381. Solid blue line; gray line marks zero. Selected term is circled.

Values at indices 1 through 20. Use the term selector and table for exact values; the line connects discrete samples and does not evaluate intermediate indices.
k = 1k = 20

Chart positions use normalized exact ratios, then truncated display coordinates. Very small differences can coincide visually; the exact table and CSV preserve them.

Positions 1–20 of 20. Full exact data is included in CSV.
PositionIndex kTermPartial sumCumulative absolute sum
11

≈ 0.5

Exact: 1/2

≈ 0.5

Exact: 1/2

≈ 0.5

Exact: 1/2

22

≈ 0.166666666667

Exact: 1/6

≈ 0.666666666667

Exact: 2/3

≈ 0.666666666667

Exact: 2/3

33

≈ 0.0833333333333

Exact: 1/12

≈ 0.75

Exact: 3/4

≈ 0.75

Exact: 3/4

44

≈ 0.05

Exact: 1/20

≈ 0.8

Exact: 4/5

≈ 0.8

Exact: 4/5

55

≈ 0.0333333333333

Exact: 1/30

≈ 0.833333333333

Exact: 5/6

≈ 0.833333333333

Exact: 5/6

66

≈ 0.0238095238095

Exact: 1/42

≈ 0.857142857143

Exact: 6/7

≈ 0.857142857143

Exact: 6/7

77

≈ 0.0178571428571

Exact: 1/56

≈ 0.875

Exact: 7/8

≈ 0.875

Exact: 7/8

88

≈ 0.0138888888889

Exact: 1/72

≈ 0.888888888889

Exact: 8/9

≈ 0.888888888889

Exact: 8/9

99

≈ 0.0111111111111

Exact: 1/90

≈ 0.9

Exact: 9/10

≈ 0.9

Exact: 9/10

1010

≈ 0.00909090909091

Exact: 1/110

≈ 0.909090909091

Exact: 10/11

≈ 0.909090909091

Exact: 10/11

1111

≈ 0.00757575757576

Exact: 1/132

≈ 0.916666666667

Exact: 11/12

≈ 0.916666666667

Exact: 11/12

1212

≈ 0.00641025641026

Exact: 1/156

≈ 0.923076923077

Exact: 12/13

≈ 0.923076923077

Exact: 12/13

1313

≈ 0.00549450549451

Exact: 1/182

≈ 0.928571428571

Exact: 13/14

≈ 0.928571428571

Exact: 13/14

1414

≈ 0.0047619047619

Exact: 1/210

≈ 0.933333333333

Exact: 14/15

≈ 0.933333333333

Exact: 14/15

1515

≈ 0.00416666666667

Exact: 1/240

≈ 0.9375

Exact: 15/16

≈ 0.9375

Exact: 15/16

1616

≈ 0.00367647058824

Exact: 1/272

≈ 0.941176470588

Exact: 16/17

≈ 0.941176470588

Exact: 16/17

1717

≈ 0.00326797385621

Exact: 1/306

≈ 0.944444444444

Exact: 17/18

≈ 0.944444444444

Exact: 17/18

1818

≈ 0.00292397660819

Exact: 1/342

≈ 0.947368421053

Exact: 18/19

≈ 0.947368421053

Exact: 18/19

1919

≈ 0.00263157894737

Exact: 1/380

≈ 0.95

Exact: 19/20

≈ 0.95

Exact: 19/20

2020

≈ 0.00238095238095

Exact: 1/420

≈ 0.952380952381

Exact: 20/21

≈ 0.952380952381

Exact: 20/21

0 zero terms. Smallest term: ≈ 0.00238095238095. Largest term: ≈ 0.5. Decimal previews use 12 significant digits; exact fractions are authoritative.

What the sum means

The upper bound is inclusive, but the selected index grid may stop below it. Arithmetic and geometric parameters advance once per selected term, independently of the numeric index step. Custom expressions evaluate k at each actual index. Division by zero, unsupported exponents or arithmetic limits produce an error for the entire requested result.

Use explicit multiplication: 2*k. Power is right-associative; -k^2 means -(k^2). No roots, trigonometric functions, symbolic infinite sums or convergence proofs for custom expressions are supplied.

OpenStax College Algebra 2e: summation notation and arithmetic/geometric identities

How to use this calculator

  1. Choose Term expression in k, Arithmetic progression or Geometric progression. Start with one of the three complete editable examples if useful.
  2. For an expression, use explicit operators such as 1/(k*(k+1)) or (-1)^k/(k+1). For a progression, enter its first selected term and common difference or ratio as a decimal or integer fraction.
  3. Enter the lower and upper index and positive step. Check the reported actual last index and term count, especially when the step does not land on the upper bound.
  4. Compare the exact sum, mean, positive/negative contributions and cancellation fraction. Treat decimal previews and chart geometry as approximate views of the exact fractions.
  5. Move the Selected term position control or term buttons to inspect a prefix, its mean and remaining finite contribution. The table follows the selected block of 20 terms.
  6. For geometric mode, inspect the continuation result and exact signed tail. Download exact CSV or use Selectable CSV for the complete dataset and original inputs.

The Formula

The mathematical logic powering this calculation is:

For positive integer step h, selected indices are k_j = lower + jh while k_j ≤ upper. N = floor((upper − lower)/h) + 1. The final selected index may be below the upper bound. Expression mode evaluates a(k_j) at each index. Partial sum S_m = a(k_0) + … + a(k_(m−1)); finite sum is S_N.
Arithmetic mode: a_j = a_0 + jd, where d applies once per selected term. S_N = N[2a_0 + (N−1)d]/2. Geometric mode: a_j = a_0 r^j. S_N = a_0(1−r^N)/(1−r) for r ≠ 1, or Na_0 for r = 1. The first term belongs to the lower selected index; changing the index step does not multiply the progression difference or ratio.
For explicit geometric mode with nonzero a_0, infinite continuation converges if |r| < 1: S_infinity = a_0/(1−r). Signed tail = S_infinity − S_N. A zero first term gives the all-zero sequence and sum zero for any rational ratio. A nonzero first term with |r| ≥ 1 has divergent infinite continuation but can still have a valid finite sum. Other modes do not assess convergence.
Positive contribution P sums positive terms; negative contribution M sums negative terms. Total = P + M. Absolute term total A = P − M. Cancellation fraction = (A − |total|)/A, defined as zero when A = 0. It describes cancellation in this finite sample, not floating-point error or infinite convergence. Mean = total/N. Prefix inspection shows S_m/m and the remaining finite contribution S_N−S_m.
Custom expressions allow k, plain decimal/integer literals, parentheses and explicit + - * / ^. Power is right-associative and binds before unary signs: -k^2 means -(k^2). Exponents must evaluate to integers from −2000 to 2000. Zero to a zero/negative power and division by zero are undefined. The full requested result is invalid if any included term is undefined; terms are not silently skipped. No functions, implicit multiplication, assignment or JavaScript execution are supported.
Bounds are integers −10000…10000, step 1…10000 and at most 2000 terms. Expressions allow 512 characters, 256 tokens and 32 nesting levels. Literals/parameters allow 40 digits including at most 20 decimal places; progression parameters also accept integer fractions with positive denominators and up to 40 digits per integer. Calculation caps are 250000 rational operations, 2048 digits per reduced numerator/denominator and 4098 temporary product digits. Exceeding a cap returns an explicit unsupported result, not a truncated answer. Exact fractions remain authoritative; decimal previews use 12 significant digits rounded half away from zero. Charts normalize exact ratios and truncate coordinates to 12 decimal places; tiny differences may coincide visually.

Worked Example

If you input:

  • Series type: Term expression in k
  • Term expression: 1/(k*(k+1))
  • Lower / upper index: 1 / 20
  • Index step: 1

Result:

The 20-term telescoping sum is exactly 20/21 (approximately 0.952380952381), because 1/[k(k+1)] = 1/k − 1/(k+1). The mean is exactly 1/21. All terms are positive, so the absolute term total is 20/21 and the cancellation fraction is zero. At the first term, the prefix is 1/2 and the remaining finite contribution is 19/42. This expression mode does not claim an infinite sum.

Frequently Asked Questions

Does sigma notation always start at one?

No. Enter any supported integer lower bound, including zero or a negative index. Expression mode substitutes each actual index into k. In a progression, the first entered term belongs to the first selected index, regardless of its numeric label.

What happens when the step misses the upper bound?

The explorer stops at the largest selected index no greater than the upper bound. For lower −5, upper 6 and step 4, it evaluates −5, −1 and 3. It does not add a term at 6 or silently change the step.

How does the index step affect arithmetic and geometric modes?

Each selected term advances the common difference or ratio once. For example, indices 2, 5 and 8 with first term 3 and difference 2 produce terms 3, 5 and 7. The step changes index labels and term count, not the meaning of the difference.

Are decimal inputs approximate?

The calculation converts entered decimals to exact rational numbers, so 0.1 is exactly 1/10. Exact fractions are retained in results and CSV. Decimal previews use 12 significant digits, and chart coordinates are approximate; those displays do not replace the exact result.

Does a leveling partial-sum chart prove convergence?

No. A finite sample is not a convergence proof. Only explicit geometric mode assesses infinite continuation using its known ratio: nonzero sequences converge for ratio magnitude below one, while an all-zero sequence converges for any ratio. Custom expressions and arithmetic mode report finite sums only.

Why is my whole calculation rejected for one undefined term?

The requested sum includes that term. Division by zero, zero to a nonpositive power or an unsupported exponent makes the expression undefined under this calculator’s rules. The error identifies the index; change the expression or index range rather than silently omitting it.

What does the cancellation fraction tell me?

It measures how much of the absolute term total disappears when positive and negative contributions offset. It is zero when contributions do not cancel and one when nonzero contributions cancel to a zero total. The all-zero series is assigned zero. It is not a floating-point error estimate.

Why are there calculation limits and paginated tables?

Exact fractions can grow very large. Explicit term, expression, digit and operation limits keep work bounded and return unsupported status when exceeded. The visible table shows 20 terms at a time; term selection and page buttons reach every row, and CSV includes all rows without display rounding. Hidden mode inputs are preserved but are not evaluated until that mode is selected.

Similar Calculators