odunbar commited on
Commit Β·
dc2991f
1
Parent(s): c29359c
change methods taxonomy
Browse files- .cache/known_methods_snapshot.json +16 -7
- CLAUDE.md +34 -13
- src/common/leaderboard.py +9 -4
- src/common/method_registry.py +33 -7
- src/data_store.py +2 -2
- src/streamlit_app.py +29 -10
.cache/known_methods_snapshot.json
CHANGED
|
@@ -3,7 +3,8 @@
|
|
| 3 |
"teki": {
|
| 4 |
"abbreviation": "TEKI",
|
| 5 |
"Method": "Tikhonov Regularized Ensemble Kalman Inversion",
|
| 6 |
-
"
|
|
|
|
| 7 |
"aliases": [
|
| 8 |
"teki"
|
| 9 |
]
|
|
@@ -11,7 +12,8 @@
|
|
| 11 |
"etki": {
|
| 12 |
"abbreviation": "ETKI",
|
| 13 |
"Method": "Ensemble Transform Kalman Inversion",
|
| 14 |
-
"
|
|
|
|
| 15 |
"aliases": [
|
| 16 |
"etki"
|
| 17 |
]
|
|
@@ -19,7 +21,8 @@
|
|
| 19 |
"iekf": {
|
| 20 |
"abbreviation": "IEKF",
|
| 21 |
"Method": "Iterative Ensemble Kalman Filter",
|
| 22 |
-
"
|
|
|
|
| 23 |
"aliases": [
|
| 24 |
"iekf",
|
| 25 |
"gnsl",
|
|
@@ -29,7 +32,8 @@
|
|
| 29 |
"uki": {
|
| 30 |
"abbreviation": "UKI",
|
| 31 |
"Method": "Unscented Kalman Inversion",
|
| 32 |
-
"
|
|
|
|
| 33 |
"aliases": [
|
| 34 |
"uki"
|
| 35 |
]
|
|
@@ -37,7 +41,8 @@
|
|
| 37 |
"abc": {
|
| 38 |
"abbreviation": "ABC",
|
| 39 |
"Method": "Approximate Bayesian Calibration",
|
| 40 |
-
"
|
|
|
|
| 41 |
"aliases": [
|
| 42 |
"abc"
|
| 43 |
]
|
|
@@ -45,7 +50,8 @@
|
|
| 45 |
"hm": {
|
| 46 |
"abbreviation": "HM",
|
| 47 |
"Method": "History Matching",
|
| 48 |
-
"
|
|
|
|
| 49 |
"aliases": [
|
| 50 |
"hm"
|
| 51 |
]
|
|
@@ -53,7 +59,8 @@
|
|
| 53 |
"ces-eki-dmc": {
|
| 54 |
"abbreviation": "CES-EKI-DMC",
|
| 55 |
"Method": "Calibrate Emulate Sample (EKI-DataMisfitController)",
|
| 56 |
-
"
|
|
|
|
| 57 |
"aliases": [
|
| 58 |
"ces-eki-dmc"
|
| 59 |
]
|
|
@@ -61,6 +68,7 @@
|
|
| 61 |
"adam": {
|
| 62 |
"abbreviation": "ADAM",
|
| 63 |
"Method": "Adaptive Moment Estimation",
|
|
|
|
| 64 |
"family": "gradient",
|
| 65 |
"aliases": [
|
| 66 |
"adam"
|
|
@@ -69,6 +77,7 @@
|
|
| 69 |
"lm": {
|
| 70 |
"abbreviation": "LM",
|
| 71 |
"Method": "Levenberg-Marquardt",
|
|
|
|
| 72 |
"family": "gradient",
|
| 73 |
"aliases": [
|
| 74 |
"lm",
|
|
|
|
| 3 |
"teki": {
|
| 4 |
"abbreviation": "TEKI",
|
| 5 |
"Method": "Tikhonov Regularized Ensemble Kalman Inversion",
|
| 6 |
+
"method_class": "optimizer",
|
| 7 |
+
"family": "kalman",
|
| 8 |
"aliases": [
|
| 9 |
"teki"
|
| 10 |
]
|
|
|
|
| 12 |
"etki": {
|
| 13 |
"abbreviation": "ETKI",
|
| 14 |
"Method": "Ensemble Transform Kalman Inversion",
|
| 15 |
+
"method_class": "optimizer",
|
| 16 |
+
"family": "kalman",
|
| 17 |
"aliases": [
|
| 18 |
"etki"
|
| 19 |
]
|
|
|
|
| 21 |
"iekf": {
|
| 22 |
"abbreviation": "IEKF",
|
| 23 |
"Method": "Iterative Ensemble Kalman Filter",
|
| 24 |
+
"method_class": "optimizer",
|
| 25 |
+
"family": "kalman",
|
| 26 |
"aliases": [
|
| 27 |
"iekf",
|
| 28 |
"gnsl",
|
|
|
|
| 32 |
"uki": {
|
| 33 |
"abbreviation": "UKI",
|
| 34 |
"Method": "Unscented Kalman Inversion",
|
| 35 |
+
"method_class": "optimizer",
|
| 36 |
+
"family": "kalman",
|
| 37 |
"aliases": [
|
| 38 |
"uki"
|
| 39 |
]
|
|
|
|
| 41 |
"abc": {
|
| 42 |
"abbreviation": "ABC",
|
| 43 |
"Method": "Approximate Bayesian Calibration",
|
| 44 |
+
"method_class": "uq",
|
| 45 |
+
"family": "uq-non-emulator",
|
| 46 |
"aliases": [
|
| 47 |
"abc"
|
| 48 |
]
|
|
|
|
| 50 |
"hm": {
|
| 51 |
"abbreviation": "HM",
|
| 52 |
"Method": "History Matching",
|
| 53 |
+
"method_class": "uq",
|
| 54 |
+
"family": "uq-emulate-within-optimize",
|
| 55 |
"aliases": [
|
| 56 |
"hm"
|
| 57 |
]
|
|
|
|
| 59 |
"ces-eki-dmc": {
|
| 60 |
"abbreviation": "CES-EKI-DMC",
|
| 61 |
"Method": "Calibrate Emulate Sample (EKI-DataMisfitController)",
|
| 62 |
+
"method_class": "uq",
|
| 63 |
+
"family": "uq-emulate-after-optimize",
|
| 64 |
"aliases": [
|
| 65 |
"ces-eki-dmc"
|
| 66 |
]
|
|
|
|
| 68 |
"adam": {
|
| 69 |
"abbreviation": "ADAM",
|
| 70 |
"Method": "Adaptive Moment Estimation",
|
| 71 |
+
"method_class": "optimizer",
|
| 72 |
"family": "gradient",
|
| 73 |
"aliases": [
|
| 74 |
"adam"
|
|
|
|
| 77 |
"lm": {
|
| 78 |
"abbreviation": "LM",
|
| 79 |
"Method": "Levenberg-Marquardt",
|
| 80 |
+
"method_class": "optimizer",
|
| 81 |
"family": "gradient",
|
| 82 |
"aliases": [
|
| 83 |
"lm",
|
CLAUDE.md
CHANGED
|
@@ -96,7 +96,7 @@ calibration_benchmark/
|
|
| 96 |
| `algorithm_type` | Canonical method key (e.g. `teki`) |
|
| 97 |
| `algorithm_alias` | Normalized raw name from NetCDF |
|
| 98 |
| `abbreviation` | Display name (e.g. `TEKI`) |
|
| 99 |
-
| `family` |
|
| 100 |
| `rmse_target` | Target RMSE level |
|
| 101 |
| `ensemble_size` | Ensemble / particle count |
|
| 102 |
| `metric` | Mean forward-model runs over `random_seed` (lower = better; β1 = failed) |
|
|
@@ -149,25 +149,46 @@ Physical dimensions of each benchmark exported from `src/data_store.py` as
|
|
| 149 |
Passed as `benchmark_dims=BENCHMARK_DIMS` to `render_leaderboard()` for column-header
|
| 150 |
annotations in the suitability table. Update if a new benchmark is added.
|
| 151 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 152 |
## Method registry (`src/common/method_registry.py`)
|
| 153 |
|
| 154 |
-
| Canonical | Abbreviation |
|
| 155 |
-
|---|---|---|---|
|
| 156 |
-
| `teki` | TEKI |
|
| 157 |
-
| `etki` | ETKI |
|
| 158 |
-
| `iekf` | IEKF |
|
| 159 |
-
| `uki` | UKI |
|
| 160 |
-
| `abc` | ABC |
|
| 161 |
-
| `hm` | HM |
|
| 162 |
-
| `ces-eki-dmc` | CES-EKI-DMC |
|
| 163 |
-
| `adam` | ADAM | gradient | β |
|
| 164 |
-
| `lm` | LM | gradient | gradient_descent |
|
| 165 |
|
| 166 |
## How to extend
|
| 167 |
|
| 168 |
### Add a method
|
| 169 |
1. Add entry to `KNOWN_METHODS` in `src/common/method_registry.py`
|
| 170 |
-
(keys: `abbreviation`, `Method`, `family`, `aliases`
|
|
|
|
| 171 |
2. Add detail-page metadata to the inline `method_meta` dict in `src/pages/MethodDetails.py`
|
| 172 |
keyed by **abbreviation** (e.g. `"TEKI"`). β οΈ Metadata is split across both files.
|
| 173 |
3. Place result NetCDF(s) in `data/` and add paths to `DATASET_FILES` in `src/data_store.py`.
|
|
|
|
| 96 |
| `algorithm_type` | Canonical method key (e.g. `teki`) |
|
| 97 |
| `algorithm_alias` | Normalized raw name from NetCDF |
|
| 98 |
| `abbreviation` | Display name (e.g. `TEKI`) |
|
| 99 |
+
| `family` | Structural sub-type β see **Method taxonomy** below |
|
| 100 |
| `rmse_target` | Target RMSE level |
|
| 101 |
| `ensemble_size` | Ensemble / particle count |
|
| 102 |
| `metric` | Mean forward-model runs over `random_seed` (lower = better; β1 = failed) |
|
|
|
|
| 149 |
Passed as `benchmark_dims=BENCHMARK_DIMS` to `render_leaderboard()` for column-header
|
| 150 |
annotations in the suitability table. Update if a new benchmark is added.
|
| 151 |
|
| 152 |
+
## Method taxonomy
|
| 153 |
+
|
| 154 |
+
Two orthogonal fields on every `KNOWN_METHODS` entry:
|
| 155 |
+
|
| 156 |
+
- **`method_class`** β `optimizer` (drives toward a single best-fit estimate) or `uq`
|
| 157 |
+
(characterizes the full posterior/uncertainty). UQ methods *can* still be scored on
|
| 158 |
+
the Optimization leaderboard β they're just usually less competitive there.
|
| 159 |
+
- **`family`** β structural mechanism, scoped within `method_class`:
|
| 160 |
+
- `optimizer` β `gradient` Β· `kalman` Β· `ensemble-other` (non-Kalman ensemble;
|
| 161 |
+
unoccupied today β reserved for e.g. CMA-ES/particle-swarm methods)
|
| 162 |
+
- `uq` β `uq-non-emulator` (samples the true forward model directly) Β·
|
| 163 |
+
`uq-emulate-after-optimize` (fits a surrogate once, after calibration) Β·
|
| 164 |
+
`uq-emulate-within-optimize` (refits a surrogate at each search iteration)
|
| 165 |
+
|
| 166 |
+
`family` deliberately avoids a "Bayesian" label β Kalman methods are approximate
|
| 167 |
+
Bayesian updates too, so the split is about mechanism (gradient vs. ensemble vs.
|
| 168 |
+
emulator timing), not philosophy. `_FAMILY_ORDER` in `src/common/leaderboard.py`
|
| 169 |
+
controls table/suitability sort order and must stay in sync with these `family` values.
|
| 170 |
+
A quick taxonomy tree is rendered on the home page (`src/streamlit_app.py`).
|
| 171 |
+
|
| 172 |
## Method registry (`src/common/method_registry.py`)
|
| 173 |
|
| 174 |
+
| Canonical | Abbreviation | method_class | family | Aliases |
|
| 175 |
+
|---|---|---|---|---|
|
| 176 |
+
| `teki` | TEKI | optimizer | kalman | β |
|
| 177 |
+
| `etki` | ETKI | optimizer | kalman | β |
|
| 178 |
+
| `iekf` | IEKF | optimizer | kalman | gnsl, gnki |
|
| 179 |
+
| `uki` | UKI | optimizer | kalman | β |
|
| 180 |
+
| `abc` | ABC | uq | uq-non-emulator | β |
|
| 181 |
+
| `hm` | HM | uq | uq-emulate-within-optimize | β |
|
| 182 |
+
| `ces-eki-dmc` | CES-EKI-DMC | uq | uq-emulate-after-optimize | β |
|
| 183 |
+
| `adam` | ADAM | optimizer | gradient | β |
|
| 184 |
+
| `lm` | LM | optimizer | gradient | gradient_descent |
|
| 185 |
|
| 186 |
## How to extend
|
| 187 |
|
| 188 |
### Add a method
|
| 189 |
1. Add entry to `KNOWN_METHODS` in `src/common/method_registry.py`
|
| 190 |
+
(keys: `abbreviation`, `Method`, `method_class`, `family`, `aliases` β see
|
| 191 |
+
**Method taxonomy** above for valid `method_class`/`family` values).
|
| 192 |
2. Add detail-page metadata to the inline `method_meta` dict in `src/pages/MethodDetails.py`
|
| 193 |
keyed by **abbreviation** (e.g. `"TEKI"`). β οΈ Metadata is split across both files.
|
| 194 |
3. Place result NetCDF(s) in `data/` and add paths to `DATASET_FILES` in `src/data_store.py`.
|
src/common/leaderboard.py
CHANGED
|
@@ -51,11 +51,16 @@ def _method_color(present_abbrevs: list[str]) -> alt.Color:
|
|
| 51 |
|
| 52 |
# Family display order in the leaderboard table (lower = earlier).
|
| 53 |
# Unknown families fall back to 99 and appear at the end.
|
|
|
|
|
|
|
|
|
|
| 54 |
_FAMILY_ORDER: dict[str, int] = {
|
| 55 |
-
"
|
| 56 |
-
"
|
| 57 |
-
"
|
| 58 |
-
"
|
|
|
|
|
|
|
| 59 |
}
|
| 60 |
|
| 61 |
|
|
|
|
| 51 |
|
| 52 |
# Family display order in the leaderboard table (lower = earlier).
|
| 53 |
# Unknown families fall back to 99 and appear at the end.
|
| 54 |
+
# Optimizer families first (gradient, kalman, ensemble-other), then UQ families
|
| 55 |
+
# ordered non-emulator -> emulate-after -> emulate-within. See method_registry.py
|
| 56 |
+
# for the full taxonomy (method_class + family).
|
| 57 |
_FAMILY_ORDER: dict[str, int] = {
|
| 58 |
+
"gradient": 0,
|
| 59 |
+
"kalman": 1,
|
| 60 |
+
"ensemble-other": 2,
|
| 61 |
+
"uq-non-emulator": 3,
|
| 62 |
+
"uq-emulate-after-optimize": 4,
|
| 63 |
+
"uq-emulate-within-optimize": 5,
|
| 64 |
}
|
| 65 |
|
| 66 |
|
src/common/method_registry.py
CHANGED
|
@@ -4,58 +4,84 @@ import json
|
|
| 4 |
from pathlib import Path
|
| 5 |
|
| 6 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 7 |
KNOWN_METHODS = {
|
| 8 |
"teki": {
|
| 9 |
"abbreviation": "TEKI",
|
| 10 |
"Method": "Tikhonov Regularized Ensemble Kalman Inversion",
|
| 11 |
-
"
|
|
|
|
| 12 |
"aliases": ["teki"],
|
| 13 |
},
|
| 14 |
"etki": {
|
| 15 |
"abbreviation": "ETKI",
|
| 16 |
"Method": "Ensemble Transform Kalman Inversion",
|
| 17 |
-
"
|
|
|
|
| 18 |
"aliases": ["etki"],
|
| 19 |
},
|
| 20 |
"iekf": {
|
| 21 |
"abbreviation": "IEKF",
|
| 22 |
"Method": "Iterative Ensemble Kalman Filter",
|
| 23 |
-
"
|
|
|
|
| 24 |
"aliases": ["iekf", "gnsl", "gnki"],
|
| 25 |
},
|
| 26 |
"uki": {
|
| 27 |
"abbreviation": "UKI",
|
| 28 |
"Method": "Unscented Kalman Inversion",
|
| 29 |
-
"
|
|
|
|
| 30 |
"aliases": ["uki"],
|
| 31 |
},
|
| 32 |
"abc": {
|
| 33 |
"abbreviation": "ABC",
|
| 34 |
"Method": "Approximate Bayesian Calibration",
|
| 35 |
-
"
|
|
|
|
| 36 |
"aliases": ["abc"],
|
| 37 |
},
|
| 38 |
"hm": {
|
| 39 |
"abbreviation": "HM",
|
| 40 |
"Method": "History Matching",
|
| 41 |
-
"
|
|
|
|
| 42 |
"aliases": ["hm"],
|
| 43 |
},
|
| 44 |
"ces-eki-dmc": {
|
| 45 |
"abbreviation": "CES-EKI-DMC",
|
| 46 |
"Method": "Calibrate Emulate Sample (EKI-DataMisfitController)",
|
| 47 |
-
"
|
|
|
|
| 48 |
"aliases": ["ces-eki-dmc"]
|
| 49 |
},
|
| 50 |
"adam": {
|
| 51 |
"abbreviation": "ADAM",
|
| 52 |
"Method": "Adaptive Moment Estimation",
|
|
|
|
| 53 |
"family": "gradient",
|
| 54 |
"aliases": ["adam"],
|
| 55 |
},
|
| 56 |
"lm": {
|
| 57 |
"abbreviation": "LM",
|
| 58 |
"Method": "Levenberg-Marquardt",
|
|
|
|
| 59 |
"family": "gradient",
|
| 60 |
"aliases": ["lm", "levenberg_marquardt", "levenberg-marquardt", "gradient_descent"],
|
| 61 |
},
|
|
|
|
| 4 |
from pathlib import Path
|
| 5 |
|
| 6 |
|
| 7 |
+
# Taxonomy
|
| 8 |
+
# --------
|
| 9 |
+
# ``method_class`` β top-level split, independent of which leaderboard a method's
|
| 10 |
+
# data happens to be registered on:
|
| 11 |
+
# "optimizer" β drives toward a single best-fit parameter estimate.
|
| 12 |
+
# "uq" β characterizes the full posterior / parameter uncertainty.
|
| 13 |
+
# UQ methods can still be scored on the Optimization leaderboard
|
| 14 |
+
# (they usually just aren't as competitive there).
|
| 15 |
+
#
|
| 16 |
+
# ``family`` β structural sub-type within a method_class:
|
| 17 |
+
# optimizer: "gradient" Β· "kalman" Β· "ensemble-other" (non-Kalman ensemble; no
|
| 18 |
+
# method occupies this slot yet, reserved for e.g. CMA-ES/particle swarm)
|
| 19 |
+
# uq: "uq-non-emulator" (posterior sampling, no surrogate model) Β·
|
| 20 |
+
# "uq-emulate-after-optimize" (build a surrogate once optimization/
|
| 21 |
+
# calibration finishes, e.g. CES) Β·
|
| 22 |
+
# "uq-emulate-within-optimize" (refit/refine a surrogate at each
|
| 23 |
+
# iteration of the search, e.g. History Matching waves)
|
| 24 |
KNOWN_METHODS = {
|
| 25 |
"teki": {
|
| 26 |
"abbreviation": "TEKI",
|
| 27 |
"Method": "Tikhonov Regularized Ensemble Kalman Inversion",
|
| 28 |
+
"method_class": "optimizer",
|
| 29 |
+
"family": "kalman",
|
| 30 |
"aliases": ["teki"],
|
| 31 |
},
|
| 32 |
"etki": {
|
| 33 |
"abbreviation": "ETKI",
|
| 34 |
"Method": "Ensemble Transform Kalman Inversion",
|
| 35 |
+
"method_class": "optimizer",
|
| 36 |
+
"family": "kalman",
|
| 37 |
"aliases": ["etki"],
|
| 38 |
},
|
| 39 |
"iekf": {
|
| 40 |
"abbreviation": "IEKF",
|
| 41 |
"Method": "Iterative Ensemble Kalman Filter",
|
| 42 |
+
"method_class": "optimizer",
|
| 43 |
+
"family": "kalman",
|
| 44 |
"aliases": ["iekf", "gnsl", "gnki"],
|
| 45 |
},
|
| 46 |
"uki": {
|
| 47 |
"abbreviation": "UKI",
|
| 48 |
"Method": "Unscented Kalman Inversion",
|
| 49 |
+
"method_class": "optimizer",
|
| 50 |
+
"family": "kalman",
|
| 51 |
"aliases": ["uki"],
|
| 52 |
},
|
| 53 |
"abc": {
|
| 54 |
"abbreviation": "ABC",
|
| 55 |
"Method": "Approximate Bayesian Calibration",
|
| 56 |
+
"method_class": "uq",
|
| 57 |
+
"family": "uq-non-emulator",
|
| 58 |
"aliases": ["abc"],
|
| 59 |
},
|
| 60 |
"hm": {
|
| 61 |
"abbreviation": "HM",
|
| 62 |
"Method": "History Matching",
|
| 63 |
+
"method_class": "uq",
|
| 64 |
+
"family": "uq-emulate-within-optimize",
|
| 65 |
"aliases": ["hm"],
|
| 66 |
},
|
| 67 |
"ces-eki-dmc": {
|
| 68 |
"abbreviation": "CES-EKI-DMC",
|
| 69 |
"Method": "Calibrate Emulate Sample (EKI-DataMisfitController)",
|
| 70 |
+
"method_class": "uq",
|
| 71 |
+
"family": "uq-emulate-after-optimize",
|
| 72 |
"aliases": ["ces-eki-dmc"]
|
| 73 |
},
|
| 74 |
"adam": {
|
| 75 |
"abbreviation": "ADAM",
|
| 76 |
"Method": "Adaptive Moment Estimation",
|
| 77 |
+
"method_class": "optimizer",
|
| 78 |
"family": "gradient",
|
| 79 |
"aliases": ["adam"],
|
| 80 |
},
|
| 81 |
"lm": {
|
| 82 |
"abbreviation": "LM",
|
| 83 |
"Method": "Levenberg-Marquardt",
|
| 84 |
+
"method_class": "optimizer",
|
| 85 |
"family": "gradient",
|
| 86 |
"aliases": ["lm", "levenberg_marquardt", "levenberg-marquardt", "gradient_descent"],
|
| 87 |
},
|
src/data_store.py
CHANGED
|
@@ -232,7 +232,7 @@ def _load_store(dataset_files: dict[str, list[str]], target_col: str, *, drop_na
|
|
| 232 |
merged["family"] = merged["algorithm_type"].map(lambda method: get_method_meta(method).get("family"))
|
| 233 |
merged["abbreviation"] = merged["abbreviation"].fillna(merged["algorithm_type"].str.upper())
|
| 234 |
merged["Method"] = merged["Method"].fillna(merged["abbreviation"])
|
| 235 |
-
merged["family"] = merged["family"].fillna("
|
| 236 |
|
| 237 |
try:
|
| 238 |
dump_method_registry_snapshot(project_root=project_root, observed_methods=set(merged["algorithm_type"].unique()))
|
|
@@ -395,6 +395,6 @@ def load_uq_budget_store() -> pd.DataFrame:
|
|
| 395 |
return get_method_meta(canonicalize_method_name(normalize_method_name(m))).get(key, fallback)
|
| 396 |
|
| 397 |
df["abbreviation"] = df["algorithm_type"].map(lambda m: _meta(m, "abbreviation", m.upper()))
|
| 398 |
-
df["family"] = df["algorithm_type"].map(lambda m: _meta(m, "family", "
|
| 399 |
df["Method"] = df["algorithm_type"].map(lambda m: _meta(m, "Method", m.upper()))
|
| 400 |
return df
|
|
|
|
| 232 |
merged["family"] = merged["algorithm_type"].map(lambda method: get_method_meta(method).get("family"))
|
| 233 |
merged["abbreviation"] = merged["abbreviation"].fillna(merged["algorithm_type"].str.upper())
|
| 234 |
merged["Method"] = merged["Method"].fillna(merged["abbreviation"])
|
| 235 |
+
merged["family"] = merged["family"].fillna("unmapped")
|
| 236 |
|
| 237 |
try:
|
| 238 |
dump_method_registry_snapshot(project_root=project_root, observed_methods=set(merged["algorithm_type"].unique()))
|
|
|
|
| 395 |
return get_method_meta(canonicalize_method_name(normalize_method_name(m))).get(key, fallback)
|
| 396 |
|
| 397 |
df["abbreviation"] = df["algorithm_type"].map(lambda m: _meta(m, "abbreviation", m.upper()))
|
| 398 |
+
df["family"] = df["algorithm_type"].map(lambda m: _meta(m, "family", "unmapped"))
|
| 399 |
df["Method"] = df["algorithm_type"].map(lambda m: _meta(m, "Method", m.upper()))
|
| 400 |
return df
|
src/streamlit_app.py
CHANGED
|
@@ -81,18 +81,37 @@ st.image(
|
|
| 81 |
width=1200,
|
| 82 |
)
|
| 83 |
|
| 84 |
-
st.subheader("Method
|
| 85 |
st.markdown(
|
| 86 |
-
"Methods
|
| 87 |
-
"citations and per-method performance charts."
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 88 |
)
|
| 89 |
-
families = {
|
| 90 |
-
"Kalman": "Ensemble Kalman variants (TEKI, ETKI, IEKF, UKI) β update an ensemble of parameter guesses via a linearized observation operator.",
|
| 91 |
-
"Bayesian": "Sampling-based approaches (ABC, HM) β explore parameter space without requiring gradient information.",
|
| 92 |
-
"Calibrate-then-emulate": "Two-stage pipelines (CES-EKI-DMC) β use an initial calibration phase to build a cheap emulator, then sample the posterior via MCMC.",
|
| 93 |
-
}
|
| 94 |
-
for family, desc in families.items():
|
| 95 |
-
st.markdown(f"- **{family}** β {desc}")
|
| 96 |
|
| 97 |
st.subheader("Key metric")
|
| 98 |
st.markdown(
|
|
|
|
| 81 |
width=1200,
|
| 82 |
)
|
| 83 |
|
| 84 |
+
st.subheader("Method taxonomy")
|
| 85 |
st.markdown(
|
| 86 |
+
"Methods split first by what they're built to do, then by how they get there. "
|
| 87 |
+
"See the **π Methods** page for citations and per-method performance charts."
|
| 88 |
+
)
|
| 89 |
+
st.markdown(
|
| 90 |
+
"""
|
| 91 |
+
- **Optimizer methods** β drive toward a single best-fit parameter estimate
|
| 92 |
+
- **Gradient-based** β `ADAM`, `LM` β follow the loss gradient (or a Gauss-Newton
|
| 93 |
+
approximation of it) directly.
|
| 94 |
+
- **Ensemble-based** β update a population of parameter guesses instead of one
|
| 95 |
+
- **Kalman** β `TEKI`, `ETKI`, `IEKF`, `UKI` β update the ensemble via a
|
| 96 |
+
(possibly linearized or unscented) Kalman step.
|
| 97 |
+
- **Non-Kalman** β *(none registered yet β e.g. genetic/CMA-ES/particle-swarm
|
| 98 |
+
methods would land here)*
|
| 99 |
+
- **UQ methods** β characterize the full posterior / parameter uncertainty, not just
|
| 100 |
+
a point estimate. Can also be scored on the Optimization leaderboard, but tend to be
|
| 101 |
+
less competitive there since they're not optimizing for speed-to-target.
|
| 102 |
+
- **Non-emulator** β `ABC` β sample the posterior directly against the true
|
| 103 |
+
forward model, no surrogate involved.
|
| 104 |
+
- **Emulate-after-optimize** β `CES-EKI-DMC` β calibrate first, then fit a
|
| 105 |
+
surrogate (e.g. a GP) once and sample the posterior through it.
|
| 106 |
+
- **Emulate-within-optimize** β `HM` β refit/refine a surrogate at each
|
| 107 |
+
iteration (wave) of the search itself.
|
| 108 |
+
"""
|
| 109 |
+
)
|
| 110 |
+
st.caption(
|
| 111 |
+
"Note: Kalman methods are Bayesian in spirit too (they're approximate Gaussian "
|
| 112 |
+
"posterior updates) β the split above is about mechanism (gradient vs. ensemble vs. "
|
| 113 |
+
"emulator timing), not whether a method is 'Bayesian'."
|
| 114 |
)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 115 |
|
| 116 |
st.subheader("Key metric")
|
| 117 |
st.markdown(
|