Model fit detail
The underlying NNLS fit behind every row on the
dashboard and its
proposed gas detail page: each (client, case)
is two independent fits — one on the zero-value transfers, one on the value
transfers — so each regime has its own slope
(→ ZERO_VALUE_TRANSFER / VALUE_TRANSFER).
No rows match this filter.
↔ Scroll horizontally to see all columns.
| Client | Case | nobs (zero / value) | R² zero | R² value | slope zero | slope CI zero | p zero | slope value | slope CI value | p value |
|---|---|---|---|---|---|---|---|---|---|---|
| besu | Contract | 418 / 418 | 0.859 | 0.924 | 0.124344 | 0.119347 – 0.129194 | 0 | 0.221274 | 0.214934 – 0.227516 | 0 |
| besu | EOA | 418 / 418 | 0.847 | 0.910 | 0.118638 | 0.1138 – 0.123614 | 0 | 0.211141 | 0.20423 – 0.218203 | 0 |
| besu | Non-existent | 418 / 418 | 0.856 | 0.932 | 0.114067 | 0.109184 – 0.119007 | 0 | 0.208206 | 0.202312 – 0.213957 | 0 |
| besu | Contract (jumpdest) | 418 / 418 | 0.840 | 0.885 | 0.144728 | 0.138908 – 0.150864 | 0 | 0.223875 | 0.216765 – 0.231497 | 0 |
| erigon | Contract | 198 / 198 | 0.963 | 0.989 | 0.139385 | 0.13564 – 0.143333 | 0 | 0.363662 | 0.358567 – 0.368637 | 0 |
| erigon | EOA | 198 / 198 | 0.940 | 0.975 | 0.09979 | 0.0961323 – 0.103471 | 0 | 0.183455 | 0.179367 – 0.187524 | 0 |
| erigon | Non-existent | 198 / 198 | 0.951 | 0.986 | 0.0965757 | 0.0937049 – 0.0995174 | 0 | 0.27437 | 0.269639 – 0.27872 | 0 |
| erigon | Contract (jumpdest) | 198 / 198 | 0.983 | 0.985 | 0.196889 | 0.193496 – 0.200485 | 0 | 0.255209 | 0.250838 – 0.259767 | 0 |
| ethrex | Contract | 275 / 275 | 0.850 | 0.991 | 0.0218686 | 0.0208882 – 0.0226894 | 0 | 0.049235 | 0.0486658 – 0.0497643 | 0 |
| ethrex | EOA | 275 / 275 | 0.826 | 0.990 | 0.0212619 | 0.0202846 – 0.0221566 | 0 | 0.0490026 | 0.0484392 – 0.0494905 | 0 |
| ethrex | Non-existent | 275 / 275 | 0.984 | 0.990 | 0.0275005 | 0.0270707 – 0.0279237 | 0 | 0.0455361 | 0.0449817 – 0.0460524 | 0 |
| ethrex | Contract (jumpdest) | 275 / 275 | 0.976 | 0.990 | 0.101629 | 0.0996386 – 0.103604 | 0 | 0.11888 | 0.117473 – 0.1202 | 0 |
| geth | Contract | 264 / 264 | 0.929 | 0.967 | 0.102104 | 0.0987778 – 0.105432 | 0 | 0.192138 | 0.188195 – 0.196262 | 0 |
| geth | EOA | 264 / 264 | 0.929 | 0.965 | 0.0865866 | 0.0835673 – 0.089723 | 0 | 0.147067 | 0.143504 – 0.150633 | 0 |
| geth | Non-existent | 264 / 264 | 0.932 | 0.964 | 0.1001 | 0.0971528 – 0.103384 | 0 | 0.183171 | 0.179013 – 0.187193 | 0 |
| geth | Contract (jumpdest) | 264 / 264 | 0.922 | 0.958 | 0.0988088 | 0.0954188 – 0.102526 | 0 | 0.145276 | 0.14187 – 0.148566 | 0 |
| nethermind | Contract | 165 / 165 | 0.959 | 0.951 | 0.0744954 | 0.072124 – 0.0767105 | 0 | 0.122448 | 0.118256 – 0.126622 | 0 |
| nethermind | EOA | 165 / 165 | 0.948 | 0.946 | 0.0684792 | 0.0663129 – 0.0708048 | 0 | 0.100045 | 0.0968498 – 0.103133 | 0 |
| nethermind | Non-existent | 165 / 165 | 0.940 | 0.931 | 0.0689176 | 0.066053 – 0.0720737 | 0 | 0.112061 | 0.107538 – 0.116252 | 0 |
| nethermind | Contract (jumpdest) | 165 / 165 | 0.981 | 0.983 | 0.143724 | 0.142588 – 0.144769 | 0 | 0.168872 | 0.167737 – 0.169957 | 0 |
p-values are bootstrap boundary tests, not conventional two-sided p-values: each is the fraction of resamples in which the coefficient collapses to zero, so a strong, clearly non-zero driver reads 0. See the methodology for details; the confidence intervals are the better guide to precision.