β campaign index Β· telemetry
| # | Candidate | FOSS implementation @ license | Paper Β· provenance | Coverage |
|---|---|---|---|---|
| #1 | l_skewness_tau3Robust, parameter-free measure of return/tick-move asymmetry built from linear combinations of order statistics (L-moments) rather than third powers β |
OpenHydrology/lmoments3 GPL-3 (lmoments3); BSD (scipy) |
DOI:10.1111/j.2517-6161.1990.t git-anchor: absent @c803ef96e + @2d308f1c |
not-covered |
| #2 | l_kurtosis_tau4Robust, parameter-free tail-peakedness measure from L-moments; concords with goodness-of-fit power and resists outliers far better than classical kurt |
OpenHydrology/lmoments3 GPL-3 (lmoments3); BSD (scipy) |
DOI:10.1111/j.2517-6161.1990.t git-anchor: absent @c803ef96e + @2d308f1c |
not-covered |
| #3 | gini_mean_differenceExpected absolute difference between two independent draws of the within-bar return/spread distribution; a robust, distribution-free scale measure equ |
cran/lmomco GPL-2.0-or-later |
DOI:10.1111/j.2517-6161.1990.t git-anchor: absent @c803ef96e + @2d308f1c |
not-covered |
| #4 | bowley_quantile_skewnessQuartile-based (Bowley/Galton) skewness coefficient: a bounded, outlier-robust asymmetry measure built from Q1, median, Q3 with fixed quartile positio |
scipy/scipy (quantiles) + ~3 lines; robustats reference impl BSD-3-Clause (scipy) |
DOI:10.2307/2987742 git-anchor: absent @c803ef96e + @2d308f1c |
not-covered |
| #5 | moors_octile_kurtosisA robust quantile-based peakedness/tail measure built from octiles, capturing how dispersed the distribution is around the +/-sigma shoulders without |
scipy/scipy BSD-3-Clause |
DOI:10.2307/2348376 git-anchor: absent @c803ef96e + @2d308f1c |
not-covered |
| #6 | medcouple_skewnessA robust skewness statistic equal to the median of a kernel of pairwise (above-median, below-median) comparisons; 25% breakdown, bounded influence, sc |
statsmodels/statsmodels BSD-3-Clause |
DOI:10.1198/106186004X12632 git-anchor: absent @c803ef96e + @2d308f1c |
not-covered |
| #7 | right_quantile_weight_tailweightA robust right-tail-heaviness descriptor: quartile skewness applied to the upper half of the distribution (above the median), capturing fat-right-tail |
cran/robustbase GPL-3.0 (robustbase); reimplementa |
DOI:10.1016/j.csda.2004.09.012 git-anchor: absent @c803ef96e + @2d308f1c |
not-covered |
| #8 | kelly_decile_skewnessA robust asymmetry measure reaching further into the tails than Bowley by using the 10th/50th/90th percentiles instead of quartiles, detecting outer-d |
numpy/numpy BSD-3-Clause |
DOI:10.1093/biomet/62.1.101 git-anchor: absent @c803ef96e + @2d308f1c |
not-covered |
| #9 | quartile_coeff_dispersion_intervals(Q3 β Q1)/(Q3 + Q1) of the within-bar inter-quote interval distribution β an outlier-robust, scale-free spread index of arrival-gap dispersion that do |
scipy/scipy BSD-3-Clause |
DOI:10.1016/j.csda.2005.05.007 git-anchor: absent @c803ef96e + @2d308f1c |
not-covered |
| #10 | sevcik_fdMaps the bar's price/quote curve into the unit square then estimates fractal dimension from the total point-to-point Euclidean curve length, capturing |
raphaelvallat/antropy BSD-3-Clause |
arXiv:1003.5266 git-anchor: absent @c803ef96e + @2d308f1c |
not-covered |
| #11 | katz_fdFractal dimension of the price path computed from the ratio of total curve length to the path's diameter (max distance from the first point), capturin |
raphaelvallat/antropy BSD-3-Clause |
DOI:10.1016/0010-4825(88)90041 git-anchor: absent @c803ef96e + @2d308f1c |
crypto-prior |
| #12 | petrosian_fdFast fractal-dimension estimate driven only by the number of sign changes in the first difference of the series, measuring path roughness/zig-zag dens |
raphaelvallat/antropy BSD-3-Clause |
DOI:10.1109/CBMS.1995.465426 git-anchor: absent @c803ef96e + @2d308f1c |
crypto-prior |
| #13 | fractal_nldNormalized Length Density: after amplitude standardization, the mean absolute first difference of the bar's series, mapped through a fixed Weierstrass |
neuropsychology/NeuroKit MIT |
DOI:10.1186/1753-4631-3-8 git-anchor: absent @c803ef96e + @2d308f1c |
not-covered |
| #14 | cid_ce_arclengthThe CID complexity estimate: the amplitude-space arc length of the series, i.e. the square root of the summed squared first differences β a single-pas |
pmontman/TSclust GPL-2.0+ |
DOI:10.1007/s10618-013-0312-3 git-anchor: absent @c803ef96e + @2d308f1c |
not-covered |
| #15 | hjorth_complexityRatio-of-ratios shape descriptor: how much the rate-of-change of the series departs from a pure sinusoid, i.e. waveform shape irregularity, derived pu |
raphaelvallat/antropy BSD-3-Clause |
DOI:10.1016/0013-4694(70)90143 git-anchor: absent @c803ef96e + @2d308f1c |
crypto-prior |
| #16 | intrabar_range_traversal_efficiencyHow directly price walked between the high and the low: the bar's high-low range divided by the total absolute tick-to-tick path length traversed insi |
raphaelvallat/antropy BSD-3-Clause |
DOI:10.1016/0010-4825(88)90041 git-anchor: absent @c803ef96e + @2d308f1c |
crypto-prior |
| #17 | mean_cosine_turning_angleDirectional persistence of the tick/bar path: the average cosine of the heading change between consecutive segments in the (time, price) plane β a sig |
JimMcL/trajr GPL-3 |
DOI:10.1016/j.jtbi.2004.03.016 git-anchor: absent @c803ef96e + @2d308f1c |
not-covered |
| #18 | vonneumann_ratioRatio of mean-square successive difference to total variance of the path β a parameter-free serial-structure / smoothness statistic; low values = tren |
AndriSignorell/DescTools GPL-2/GPL-3 |
DOI:10.1214/aoms/1177731677 git-anchor: absent @c803ef96e + @2d308f1c |
not-covered |
| #19 | bartels_rank_vn_ratioRank version of the von Neumann ratio: mean-square successive rank-difference over rank variance; a distribution-free serial-dependence/randomness sta |
cran/randtests GPL-2.0-or-later |
DOI:10.1080/01621459.1982.1047 git-anchor: absent @c803ef96e + @2d308f1c |
not-covered |
| #20 | mann_kendall_s_zMann-Kendall monotonic-trend statistic: the normalized sum over all ordered bar pairs of the sign of (later value minus earlier value) β a rank-based, |
cran/Kendall GPL-2 (Kendall pkg); pymannkendall |
DOI:10.2307/1907187 git-anchor: absent @c803ef96e + @2d308f1c |
not-covered |
| #21 | turning_point_rate_zStandardized count of turning points (bars that are a strict local max or local min vs their two neighbors) β high values indicate oscillatory/mean-re |
cran/randtests GPL (>= 2) |
DOI:10.1111/j.2517-6161.1954.t git-anchor: absent @c803ef96e + @2d308f1c |
not-covered |
| #22 | difference_sign_test_zStandardized count of positive first-differences (number of up-moves) across the lookback β a parameter-free sign-balance-of-increments statistic dist |
cran/aTSA GPL (>= 2) |
DOI:10.1093/biomet/42.1-2.80 git-anchor: absent @c803ef96e + @2d308f1c |
not-covered |
| #23 | cox_stuart_trend_zCox-Stuart quick sign test for trend: pair the i-th value of the first half with the i-th of the second half, count how often the second exceeds the f |
cran/randtests GPL (>= 2) |
DOI:10.1093/biomet/42.1-2.80 git-anchor: absent @c803ef96e + @2d308f1c |
not-covered |
| #24 | foster_stuart_records_dFoster-Stuart record-breaking trend statistic: difference between counts of upper-record and lower-record breaks over the lookback β distribution-free |
JorgeCastilloMateo/RecordTest GPL (>= 3) |
DOI:10.1111/j.2517-6161.1954.t git-anchor: absent @c803ef96e + @2d308f1c |
not-covered |
| #25 | pomeau_irreversibility_lag1The lag-1 time-asymmetric cubic statistic E[x_t^2 * x_{t-1} - x_t * x_{t-1}^2], which is positive/negative when the series rises and falls with differ |
MZanin/irreversibilitytestslibrary unknown/academic (verify on GitLab |
arXiv:0708.4022 git-anchor: absent @c803ef96e + @2d308f1c |
not-covered |
| #26 | ehlers_increment_asymmetryEhlers' time-irreversibility index: the normalized third moment (skewness) of the first-difference series of close prices; zero means time-reversible, |
MZanin/irreversibilitytestslibrary unknown/academic (verify on GitLab |
DOI:10.3390/e23111474 git-anchor: absent @c803ef96e + @2d308f1c |
not-covered |
| #27 | porta_increment_sign_asymmetryPorta's index: the fraction of negative first-differences among all nonzero first-differences; a value far from 0.5 indicates acceleration/deceleratio |
MZanin/irreversibilitytestslibrary unknown/academic (verify on GitLab |
DOI:10.3390/e23111474 git-anchor: absent @c803ef96e + @2d308f1c |
not-covered |
| #28 | guzik_increment_magnitude_asymmetryGuzik's index: the share of total squared-increment 'energy' that comes from positive increments; differs from Porta by weighting by magnitude rather |
MZanin/irreversibilitytestslibrary unknown/academic (verify on GitLab |
DOI:10.3390/e23111474 git-anchor: absent @c803ef96e + @2d308f1c |
not-covered |
| #29 | costa_accel_decel_asymmetryCosta-Goldberger-Peng time-irreversibility index: difference between the rates/energies of consecutive increases vs decreases ('accelerations' vs 'dec |
MZanin/irreversibilitytestslibrary unknown/academic (verify on GitLab |
DOI:10.1103/PhysRevLett.95.198 git-anchor: absent @c803ef96e + @2d308f1c |
not-covered |
| #30 | dhvg_indegree_outdegree_kldLacasa directed-horizontal-visibility-graph irreversibility: map the price series to a directed horizontal visibility graph and measure the Kullback-L |
CarlosBergillos/ts2vg MIT (ts2vg) |
arXiv:1108.1691 git-anchor: absent @c803ef96e + @2d308f1c |
not-covered |
| #31 | trading_polarityNet buy-vs-sell trade-count imbalance within the bar, signed and bounded in [-1,+1], using counts of buyer- and seller-initiated ticks rather than the β forex-roadmap-mentioned (listed as an UNTESTED inventory item in the May Gate extended-inventory; not yet evaluated β ve π FOREX-MENTIONED-ONLY (roadmap inventory, not evaluated) |
reimplement from paper n/a |
DOI:10.3390/e22080897 git-anchor: absent @c803ef96e + @2d308f1c |
not-covered |
| #32 | quote_revision_direction_asymmetryImbalance between the count of upward and downward revisions of the best-bid (and symmetrically best-ask) quote across the intra-bar quote stream β ca π FOREX-COVERED: evaluated on forex_bars in the 2026-05-26/27 Gate campaign |
reimplement from paper n/a |
DOI:10.1093/rfs/10.4.1035 git-anchor: absent @c803ef96e + @2d308f1c |
β forex-dup |
| #33 | quote_to_trade_ratioRatio of the number of quote updates to the number of executed trades within the bar β quoting intensity relative to execution, a market-quality/HFT-a |
reimplement from paper n/a |
DOI:10.1016/j.jbankfin.2014.09 git-anchor: absent @c803ef96e + @2d308f1c |
not-covered |
| #34 | buy_sell_duration_asymmetryAsymmetry between mean inter-trade waiting times following buyer-initiated ticks versus seller-initiated ticks within the bar β a signed-time intrinsi |
reimplement from paper n/a |
arXiv:physics/0608273 git-anchor: absent @c803ef96e + @2d308f1c |
not-covered |
| #35 | signed_volume_gini_concentrationGini concentration of per-tick signed-volume contributions within the bar β measures whether directional order flow is dispersed across many ticks or |
reimplement from paper n/a |
DOI:10.3390/jrfm17080325 git-anchor: absent @c803ef96e + @2d308f1c |
not-covered |
| #36 | tick_rule_effective_spread_capture_ratioMean signed (price - mid) over the bar's ticks expressed as a fraction of the mean half-spread β a parameter-free proxy for how much of the quoted spr |
mgao6767/frds MIT (unverified β github/pypi fetc |
DOI:10.1111/j.1540-6261.1991.t git-anchor: absent @c803ef96e + @2d308f1c |
not-covered |
| #37 | cumulative_residual_entropyRao's CDF-based entropy: CRE = -β« S(x) log S(x) dx, replacing the density in Shannon entropy with the survival function; a parameter-free uncertainty |
reimplement from paper none |
DOI:10.1109/TIT.2004.828057 git-anchor: absent @c803ef96e + @2d308f1c |
not-covered |
| #38 | cumulative_residual_extropyThe CDF-based dual of extropy: CREx = -1/2 β« S(x)^2 dx where S is the survival function; a knob-free concentration/dispersion measure that replaces th |
reimplement from paper none |
DOI:10.1017/S0269964819000196 git-anchor: absent @c803ef96e + @2d308f1c |
not-covered |
| #39 | empirical_cumulative_entropyDi CrescenzoβLongobardi cumulative entropy: CE = -β« F(x) log F(x) dx using the CDF (left-tail dual of Rao's CRE); a parameter-free spread/uncertainty |
reimplement from paper none |
DOI:10.1016/j.jspi.2009.05.038 git-anchor: absent @c803ef96e + @2d308f1c |
not-covered |
| #40 | lziv_complexity_signsLempel-Ziv (1976) complexity of the binary up/down sign sequence of the path: counts distinct sub-patterns produced by recursive copy, an entropy-rate π CRYPTO-PRIOR: present in patterns/crypto lit-scan (provenance only; still forex-proposable |
raphaelvallat/antropy BSD-3-Clause |
DOI:10.1109/TIT.1976.1055501 git-anchor: absent @c803ef96e + @2d308f1c |
crypto-prior |
| #41 | interval_entropy_dispersion_coeffEntropy-based (information) dispersion coefficient of the unit-mean-normalized inter-quote interval distribution β quantifies arrival-timing randomnes |
gregversteeg/NPEET MIT (NPEET); BSD-3-Clause (scipy) |
DOI:10.1016/j.ins.2012.12.046 git-anchor: absent @c803ef96e + @2d308f1c |
not-covered |
| #42 | candle_signed_shoelace_areaSigned (oriented) area enclosed by treating one bar as the closed 4-vertex polyline openβhighβlowβcloseβopen in (ordinal-position, price) space, norma |
shapely/shapely BSD-3-Clause |
no-DOI β shoelace / surveyor's-area form git-anchor: absent @c803ef96e + @2d308f1c |
not-covered |
| #43 | bid_ask_close_location_divergenceDifference between where the close sits within the BID candle's range versus within the ASK candle's range β a spread-geometry asymmetry that mid-pric |
twopirllc/pandas-ta MIT |
no-DOI β composed from Williams' Close-L git-anchor: absent @c803ef96e + @2d308f1c |
not-covered |
| #44 | ohlc_ordinal_pattern_indexCategorical index naming which of the admissible orderings the four OHLC values realize within the single bar (since H is max and L is min, this resol |
arsenetar/ordpy MIT |
DOI:10.1103/PhysRevLett.88.174 git-anchor: absent @c803ef96e + @2d308f1c |
crypto-prior |
| #45 | heikin_ashi_open_gap_signParameter-free Heikin-Ashi-derived single-bar invariant: the sign/normalized gap between the bar's HA-open ( (prev O + prev C)/2 ) and the bar's true |
twopirllc/pandas-ta MIT |
no-DOI β Valcu, 'Using the Heikin-Ashi T git-anchor: absent @c803ef96e + @2d308f1c |
not-covered |
| #46 | intrabar_extreme_order_signA single sign indicating whether the bar's high or its low occurred first in time within the bar. |
reimplement from paper none |
arXiv:2509.16137 git-anchor: absent @c803ef96e + @2d308f1c |
not-covered |
| #47 | local_variation_lvMean of 3(Ο_i β Ο_{i+1})Β² / (Ο_i + Ο_{i+1})Β² over consecutive inter-quote intervals β a rate-robust local irregularity index where Poisson timing give |
NeuralEnsemble/elephant BSD-3-Clause (Modified BSD) |
DOI:10.1162/089976603322518759 git-anchor: absent @c803ef96e + @2d308f1c |
not-covered |
| #48 | cv2_consecutive_intervalsMean of 2|Ο_{i+1} β Ο_i| / (Ο_{i+1} + Ο_i) over consecutive inter-quote intervals β a firing-rate-robust local timing-variability index (Poisson = 1). |
NeuralEnsemble/elephant BSD-3-Clause (Modified BSD) |
DOI:10.1152/jn.1996.75.5.1806 git-anchor: absent @c803ef96e + @2d308f1c |
not-covered |
| #49 | gini_interquote_intervalGini concentration coefficient of the within-bar inter-quote interval lengths β measures how unequally arrival gaps are distributed (Poisson/exponenti |
AndriSignorell/DescTools GPL-2.0-or-later (DescTools); triv |
arXiv:physics/0701008 git-anchor: absent @c803ef96e + @2d308f1c |
not-covered |
| #50 | rogers_satchell_var_barDrift-independent single-bar variance built from high/low/close relative to the bar open, so it is unbiased even when the bar trends. π CRYPTO-PRIOR: present in patterns/crypto lit-scan (provenance only; still forex-proposable |
cran/TTR GPL-2/3 |
DOI:10.1214/aoap/1177005835 git-anchor: absent @c803ef96e + @2d308f1c |
crypto-prior |
| #51 | overnight_gap_logret_sqSquared log gap between this bar's open and the previous bar's close β the pure close-to-open (overnight/inter-bar) variance component. π CRYPTO-PRIOR: present in patterns/crypto lit-scan (provenance only; still forex-proposable |
jasonstrimpel/volatility-trading MIT |
DOI:10.1086/209650 git-anchor: absent @c803ef96e + @2d308f1c |
crypto-prior |
| #52 | edge_spread_bpsEfficient Discrete Generalized Estimator: a single GMM-based effective-spread estimate that combines all four OHLC prices, asymptotically unbiased und |
eguidotti/bidask MIT |
DOI:10.1016/j.jfineco.2024.103 git-anchor: absent @c803ef96e + @2d308f1c |
not-covered |
| #53 | ohlc_generalized_spread_bpsA specific generalized OHLC-combination spread estimator (e.g., OHLC.CHLO or OHL.HLC average) from the same Ardia-Guidotti-Kroencke family, distinct f |
eguidotti/bidask MIT |
DOI:10.1016/j.jfineco.2024.103 git-anchor: absent @c803ef96e + @2d308f1c |
not-covered |