diff --git a/examples/pilnu_model.json b/examples/pilnu_model.json
index f4406ca..a3c82ae 100644
--- a/examples/pilnu_model.json
+++ b/examples/pilnu_model.json
@@ -1,327 +1,3 @@
-{
- "spec": {
- "channels": [
- {
- "name": "singlechannel",
- "samples": [
- {
- "name": "signal",
- "data": [
- 2538.382092422792,
- 5425.385416572461,
- 6453.507789727192,
- 5974.580002549716,
- 4727.449601010529,
- 2667.139613303666,
- 224.270383143203
- ],
- "modifiers": [
- {
- "name": "mu",
- "type": "normfactor",
- "data": null
- },
- {
- "name": "stat",
- "type": "staterror",
- "data": [
- 50.3823589406331,
- 73.65721564498932,
- 80.33372759761116,
- 77.29540738329617,
- 68.75645134102348,
- 51.64435703253228,
- 14.975659689749998
- ]
- },
- {
- "name": "theory",
- "type": "custom",
- "data": {
- "expr": "weight_function",
- "ibin": [
- 0,
- 1,
- 2,
- 3,
- 4,
- 5,
- 6
- ]
- }
- }
- ]
- }
- ]
- }
- ]
- },
- "new_pars": {
- "cvl": {
- "inits": [
- 1.0
- ],
- "bounds": [
- [
- -2.0,
- 2.0
- ]
- ],
- "paramset_type": "unconstrained"
- },
- "csl": {
- "inits": [
- 0.0
- ],
- "bounds": [
- [
- -2.0,
- 2.0
- ]
- ],
- "paramset_type": "unconstrained"
- },
- "ct": {
- "inits": [
- 0.0
- ],
- "bounds": [
- [
- -2.0,
- 2.0
- ]
- ],
- "paramset_type": "unconstrained"
- },
- "FF": {
- "inits": [
- 0.6570195222647179,
- -2.73517250970471,
- 3.8484671014044984,
- 0.019119881751032075,
- -2.3528703504189674
- ],
- "bounds": [],
- "cov": [
- [
- 0.00014896007219431363,
- 0.0016165484172655695,
- 0.0009673651261666283,
- 0.00271486990627058,
- 0.011957494655755475
- ],
- [
- 0.0016165484172655695,
- 0.033202621553503606,
- 0.14610410621238415,
- 0.038434958583196156,
- 0.20447965388564962
- ],
- [
- 0.0009673651261666283,
- 0.14610410621238415,
- 1.7120597769326753,
- 0.07598171709130717,
- 0.8459425323071735
- ],
- [
- 0.00271486990627058,
- 0.038434958583196156,
- 0.07598171709130717,
- 0.060393631388842614,
- 0.30510741422357107
- ],
- [
- 0.011957494655755475,
- 0.20447965388564962,
- 0.8459425323071735,
- 0.30510741422357107,
- 2.1291211607836567
- ]
- ],
- "paramset_type": "constrained_by_normal"
- }
- },
- "map": [
- [
- 904.8299803133875,
- 796.3891868913612,
- 537.8663353732505,
- 222.08674492830988,
- 28.628369463414945,
- 26.025790421286313,
- 16.483000600148,
- 6.072684431633474,
- 0.0,
- 0.0,
- 0.0,
- 0.0,
- 0.0,
- 0.0,
- 0.0,
- 0.0,
- 0.0,
- 0.0,
- 0.0,
- 0.0
- ],
- [
- 634.0087936045725,
- 573.6270037374703,
- 444.70372266987386,
- 158.29820586780832,
- 1192.9483349419368,
- 1191.3163946752584,
- 881.2477440063556,
- 347.6032768025069,
- 0.0,
- 0.0,
- 0.8159701333392181,
- 0.8159701333392181,
- 0.0,
- 0.0,
- 0.0,
- 0.0,
- 0.0,
- 0.0,
- 0.0,
- 0.0
- ],
- [
- 3.7217461301771584,
- 5.210444582248021,
- 3.7217461301771584,
- 1.4886984520708633,
- 1543.780294797485,
- 1598.1177882980717,
- 1178.3048248140883,
- 482.33829847095967,
- 462.9852185940385,
- 569.4271579171052,
- 439.9103925869401,
- 164.5011789538304,
- 0.0,
- 0.0,
- 0.0,
- 0.0,
- 0.0,
- 0.0,
- 0.0,
- 0.0
- ],
- [
- 0.0,
- 0.0,
- 0.0,
- 0.0,
- 92.21424073012513,
- 87.05540208788037,
- 72.22374099142668,
- 23.21477389010143,
- 1601.174543586718,
- 1803.0141054645444,
- 1478.652125833405,
- 542.9677670962612,
- 87.70025691816096,
- 79.31714412451322,
- 74.15830548226846,
- 32.88759634431036,
- 0.0,
- 0.0,
- 0.0,
- 0.0
- ],
- [
- 0.0,
- 0.0,
- 0.0,
- 0.0,
- 0.0,
- 0.0,
- 0.0,
- 0.0,
- 344.0079604636319,
- 380.99261600685,
- 310.56977873962643,
- 141.85895276850798,
- 1017.8379861140448,
- 1189.5886467873456,
- 927.6562232826361,
- 404.29801539024777,
- 3.039834702182314,
- 5.066391170303857,
- 1.519917351091157,
- 1.0132782340607713
- ],
- [
- 0.0,
- 0.0,
- 0.0,
- 0.0,
- 0.0,
- 0.0,
- 0.0,
- 0.0,
- 0.31463248947784195,
- 0.6292649789556839,
- 0.31463248947784195,
- 1.8877949368670517,
- 477.29748653788624,
- 561.9336262074257,
- 455.902477253393,
- 195.072143476262,
- 256.1108464349633,
- 324.0714641621772,
- 267.43761605616567,
- 126.16762828061462
- ],
- [
- 0.0,
- 0.0,
- 0.0,
- 0.0,
- 0.0,
- 0.0,
- 0.0,
- 0.0,
- 0.0,
- 0.0,
- 0.0,
- 0.0,
- 1.1017126895116764,
- 1.6286187584085652,
- 1.4370165515369693,
- 0.5269060688968887,
- 64.23463985370253,
- 67.06077240505857,
- 61.26480564719279,
- 27.015911168895023
- ]
- ],
- "bins": [
- [
- 3.4,
- 8.002,
- 12.604000000000001,
- 17.206,
- 21.808,
- 26.41
- ],
- [
- -1.0,
- -0.5,
- 0.0,
- 0.5,
- 1.0
- ]
- ],
- "data": [
- 6106.517959181146,
- 14631.976431038858,
- 19258.547525214723,
- 19815.744079692413,
- 17167.973119691647,
- 10369.02832323176,
- 1050.0279942080635
- ]
-}
\ No newline at end of file
+version https://git-lfs.github.com/spec/v1
+oid sha256:780e818517422e2ad22823fd334a292b33274075a3e93dc456bd7c3ea4829ecf
+size 8683
diff --git a/examples/pilnu_sample.ipynb b/examples/pilnu_sample.ipynb
index 7ae3d24..a6c342b 100644
--- a/examples/pilnu_sample.ipynb
+++ b/examples/pilnu_sample.ipynb
@@ -2,9 +2,17 @@
"cells": [
{
"cell_type": "code",
- "execution_count": 2,
+ "execution_count": 1,
"metadata": {},
- "outputs": [],
+ "outputs": [
+ {
+ "name": "stderr",
+ "output_type": "stream",
+ "text": [
+ "WARNING (pytensor.tensor.blas): Using NumPy C-API based implementation for BLAS functions.\n"
+ ]
+ }
+ ],
"source": [
"%reload_ext autoreload\n",
"%autoreload 2\n",
@@ -21,22 +29,9 @@
},
{
"cell_type": "code",
- "execution_count": 3,
+ "execution_count": 2,
"metadata": {},
- "outputs": [
- {
- "ename": "AttributeError",
- "evalue": "module 'eos' has no attribute 'Parameters'",
- "output_type": "error",
- "traceback": [
- "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
- "\u001b[0;31mAttributeError\u001b[0m Traceback (most recent call last)",
- "\u001b[1;32m/home/l/Lorenz.Gaertner/redist/examples/pilnu_sample.ipynb Cell 2\u001b[0m line \u001b[0;36m1\n\u001b[0;32m----> 1\u001b[0m null \u001b[39m=\u001b[39m pilnu_utils\u001b[39m.\u001b[39;49mnull_pred()\n\u001b[1;32m 2\u001b[0m alt \u001b[39m=\u001b[39m pilnu_utils\u001b[39m.\u001b[39malt_pred()\n",
- "File \u001b[0;32m/filer/z-sv-pool12c/l/Lorenz.Gaertner/redist/examples/pilnu_utils.py:36\u001b[0m, in \u001b[0;36mnull_pred.__init__\u001b[0;34m(self)\u001b[0m\n\u001b[1;32m 35\u001b[0m \u001b[39mdef\u001b[39;00m \u001b[39m__init__\u001b[39m(\u001b[39mself\u001b[39m):\n\u001b[0;32m---> 36\u001b[0m p \u001b[39m=\u001b[39m eos\u001b[39m.\u001b[39;49mParameters()\n\u001b[1;32m 37\u001b[0m o \u001b[39m=\u001b[39m eos\u001b[39m.\u001b[39mOptions({\u001b[39m'\u001b[39m\u001b[39mform-factors\u001b[39m\u001b[39m'\u001b[39m: \u001b[39m'\u001b[39m\u001b[39mBSZ2015\u001b[39m\u001b[39m'\u001b[39m, \u001b[39m'\u001b[39m\u001b[39ml\u001b[39m\u001b[39m'\u001b[39m: \u001b[39m'\u001b[39m\u001b[39mtau\u001b[39m\u001b[39m'\u001b[39m, \u001b[39m'\u001b[39m\u001b[39mmodel\u001b[39m\u001b[39m'\u001b[39m:\u001b[39m'\u001b[39m\u001b[39mWET\u001b[39m\u001b[39m'\u001b[39m})\n\u001b[1;32m 38\u001b[0m k \u001b[39m=\u001b[39m eos\u001b[39m.\u001b[39mKinematics({\u001b[39m'\u001b[39m\u001b[39mq2\u001b[39m\u001b[39m'\u001b[39m: \u001b[39m5.0\u001b[39m, \u001b[39m'\u001b[39m\u001b[39mcos(theta_l)\u001b[39m\u001b[39m'\u001b[39m: \u001b[39m0.0\u001b[39m,})\n",
- "\u001b[0;31mAttributeError\u001b[0m: module 'eos' has no attribute 'Parameters'"
- ]
- }
- ],
+ "outputs": [],
"source": [
"null = pilnu_utils.null_pred()\n",
"alt = pilnu_utils.alt_pred()"
@@ -44,11 +39,11 @@
},
{
"cell_type": "code",
- "execution_count": 3,
+ "execution_count": 4,
"metadata": {},
"outputs": [],
"source": [
- "model, alt_yields = modifier.load('knunu_model.json', alt.distribution, null.distribution, return_data=True)"
+ "model, alt_yields = modifier.load('pilnu_model.json', alt.distribution, null.distribution, return_data=True)"
]
},
{
@@ -469,167 +464,167 @@
" FF_decorrelated[3] (chain, draw, FF_decorrelated[3]_dim_0) float64 ...\n",
" FF_decorrelated[4] (chain, draw, FF_decorrelated[4]_dim_0) float64 ...\n",
" mu (chain, draw, mu_dim_0) float64 1.0 1.0 ... 1.0\n",
- " stat (chain, draw, stat_dim_0) float64 1.109 ... 1.089\n",
- " cvl (chain, draw, cvl_dim_0) float64 0.2203 ... 0.2313\n",
- " csl (chain, draw, csl_dim_0) float64 0.779 ... 0.7782\n",
- " ct (chain, draw, ct_dim_0) float64 0.03225 ... 0.0...\n",
+ " stat (chain, draw, stat_dim_0) float64 1.026 ... 1.01\n",
+ " cvl (chain, draw, cvl_dim_0) float64 0.7268 ... 0.9691\n",
+ " csl (chain, draw, csl_dim_0) float64 0.508 ... 0.5301\n",
+ " ct (chain, draw, ct_dim_0) float64 0.7451 ... 0.498\n",
"Attributes:\n",
- " created_at: 2023-11-24T19:48:21.837970\n",
+ " created_at: 2023-11-27T17:35:25.353355\n",
" arviz_version: 0.16.1\n",
" inference_library: pymc\n",
- " inference_library_version: 5.9.2\n",
- " sampling_time: 8654.44973897934\n",
- " tuning_steps: 1500
- chain: 4
- draw: 10000
- FF_decorrelated[0]_dim_0: 1
- FF_decorrelated[1]_dim_0: 1
- FF_decorrelated[2]_dim_0: 1
- FF_decorrelated[3]_dim_0: 1
- FF_decorrelated[4]_dim_0: 1
- mu_dim_0: 1
- stat_dim_0: 7
- cvl_dim_0: 1
- csl_dim_0: 1
- ct_dim_0: 1
chain
(chain)
int64
0 1 2 3
draw
(draw)
int64
0 1 2 3 4 ... 9996 9997 9998 9999
array([ 0, 1, 2, ..., 9997, 9998, 9999])
FF_decorrelated[0]_dim_0
(FF_decorrelated[0]_dim_0)
int64
0
FF_decorrelated[1]_dim_0
(FF_decorrelated[1]_dim_0)
int64
0
FF_decorrelated[2]_dim_0
(FF_decorrelated[2]_dim_0)
int64
0
FF_decorrelated[3]_dim_0
(FF_decorrelated[3]_dim_0)
int64
0
FF_decorrelated[4]_dim_0
(FF_decorrelated[4]_dim_0)
int64
0
mu_dim_0
(mu_dim_0)
int64
0
stat_dim_0
(stat_dim_0)
int64
0 1 2 3 4 5 6
array([0, 1, 2, 3, 4, 5, 6])
cvl_dim_0
(cvl_dim_0)
int64
0
csl_dim_0
(csl_dim_0)
int64
0
ct_dim_0
(ct_dim_0)
int64
0
FF_decorrelated[0]
(chain, draw, FF_decorrelated[0]_dim_0)
float64
0.4126 0.4126 ... 0.008686 -0.6452
array([[[ 0.41256776],\n",
- " [ 0.41256776],\n",
- " [ 0.41256776],\n",
- " ...,\n",
- " [ 0.01251961],\n",
- " [ 1.3082441 ],\n",
- " [ 0.97616677]],\n",
- "\n",
- " [[-0.36704887],\n",
- " [-0.36704887],\n",
- " [-0.36704887],\n",
- " ...,\n",
- " [-0.97837711],\n",
- " [-0.97837711],\n",
- " [-0.97837711]],\n",
- "\n",
- " [[ 1.17796078],\n",
- " [ 0.89771861],\n",
- " [ 0.95475332],\n",
- " ...,\n",
- " [ 0.49359111],\n",
- " [ 0.49359111],\n",
- " [ 0.51151981]],\n",
- "\n",
- " [[ 0.59841519],\n",
- " [ 0.59841519],\n",
- " [-0.01904167],\n",
- " ...,\n",
- " [ 0.00868571],\n",
- " [ 0.00868571],\n",
- " [-0.64518705]]])
FF_decorrelated[1]
(chain, draw, FF_decorrelated[1]_dim_0)
float64
0.8715 0.8715 ... 0.1297 0.02299
array([[[ 0.87148849],\n",
- " [ 0.87148849],\n",
- " [ 0.87148849],\n",
- " ...,\n",
- " [ 0.14491692],\n",
- " [-0.20216996],\n",
- " [-0.20216996]],\n",
- "\n",
- " [[ 1.70857927],\n",
- " [ 1.70857927],\n",
- " [ 1.70857927],\n",
- " ...,\n",
- " [ 0.82145925],\n",
- " [ 0.82145925],\n",
- " [ 1.12600751]],\n",
- "\n",
- " [[-0.65821828],\n",
- " [ 1.24855439],\n",
- " [ 1.84304703],\n",
- " ...,\n",
- " [ 0.08044903],\n",
- " [ 0.82988036],\n",
- " [ 1.55872938]],\n",
- "\n",
- " [[ 0.33988554],\n",
- " [ 0.33988554],\n",
- " [-0.83678201],\n",
- " ...,\n",
- " [ 0.12966171],\n",
- " [ 0.12966171],\n",
- " [ 0.02299128]]])
FF_decorrelated[2]
(chain, draw, FF_decorrelated[2]_dim_0)
float64
-0.9215 0.6733 ... 0.4137 0.4137
array([[[-0.92149157],\n",
- " [ 0.67331855],\n",
- " [ 0.67331855],\n",
- " ...,\n",
- " [-1.31862589],\n",
- " [-1.09327228],\n",
- " [-1.09327228]],\n",
- "\n",
- " [[-0.35857489],\n",
- " [ 0.882324 ],\n",
- " [ 0.882324 ],\n",
- " ...,\n",
- " [-1.41365216],\n",
- " [-0.66184362],\n",
- " [ 0.20250452]],\n",
- "\n",
- " [[ 1.80549966],\n",
- " [ 2.19517895],\n",
- " [ 0.13224216],\n",
- " ...,\n",
- " [ 1.17089685],\n",
- " [-0.45396506],\n",
- " [-0.45396506]],\n",
- "\n",
- " [[-2.00589164],\n",
- " [-2.00589164],\n",
- " [-0.57601023],\n",
- " ...,\n",
- " [ 0.41372485],\n",
- " [ 0.41372485],\n",
- " [ 0.41372485]]])
FF_decorrelated[3]
(chain, draw, FF_decorrelated[3]_dim_0)
float64
0.5074 -1.442 ... -0.8768 1.173
array([[[ 0.50737888],\n",
- " [-1.44152552],\n",
- " [-1.44152552],\n",
- " ...,\n",
- " [-0.4118772 ],\n",
- " [-0.03262187],\n",
- " [-0.78576106]],\n",
- "\n",
- " [[-1.37603415],\n",
- " [ 0.9163817 ],\n",
- " [ 0.9163817 ],\n",
- " ...,\n",
- " [-0.97237879],\n",
- " [-1.35325822],\n",
- " [-1.35325822]],\n",
- "\n",
- " [[-0.81890771],\n",
- " [-0.81890771],\n",
- " [-0.81890771],\n",
- " ...,\n",
- " [-1.43335646],\n",
- " [-1.43335646],\n",
- " [ 1.07743062]],\n",
- "\n",
- " [[-0.14262192],\n",
- " [-0.14262192],\n",
- " [-0.14262192],\n",
- " ...,\n",
- " [-0.26554058],\n",
- " [-0.87680095],\n",
- " [ 1.17256664]]])
FF_decorrelated[4]
(chain, draw, FF_decorrelated[4]_dim_0)
float64
0.1091 -0.8406 ... -0.0105 0.6774
array([[[ 0.10907296],\n",
- " [-0.84060627],\n",
- " [-0.84060627],\n",
- " ...,\n",
- " [-0.64851146],\n",
- " [-0.64851146],\n",
- " [-0.71705656]],\n",
- "\n",
- " [[-0.15698168],\n",
- " [-0.15698168],\n",
- " [-0.15698168],\n",
- " ...,\n",
- " [ 1.13880639],\n",
- " [ 0.12227332],\n",
- " [ 0.31911036]],\n",
- "\n",
- " [[-0.51030332],\n",
- " [-0.09820725],\n",
- " [-0.09820725],\n",
- " ...,\n",
- " [-1.22424222],\n",
- " [-1.27041067],\n",
- " [-0.87259819]],\n",
- "\n",
- " [[ 0.49714595],\n",
- " [ 0.49714595],\n",
- " [ 0.49714595],\n",
- " ...,\n",
- " [-1.36546575],\n",
- " [-0.0104998 ],\n",
- " [ 0.67738229]]])
mu
(chain, draw, mu_dim_0)
float64
1.0 1.0 1.0 1.0 ... 1.0 1.0 1.0 1.0
array([[[1.],\n",
+ " inference_library_version: 5.10.0\n",
+ " sampling_time: 6302.489182949066\n",
+ " tuning_steps: 1500
- chain: 4
- draw: 10000
- FF_decorrelated[0]_dim_0: 1
- FF_decorrelated[1]_dim_0: 1
- FF_decorrelated[2]_dim_0: 1
- FF_decorrelated[3]_dim_0: 1
- FF_decorrelated[4]_dim_0: 1
- mu_dim_0: 1
- stat_dim_0: 7
- cvl_dim_0: 1
- csl_dim_0: 1
- ct_dim_0: 1
chain
(chain)
int64
0 1 2 3
draw
(draw)
int64
0 1 2 3 4 ... 9996 9997 9998 9999
array([ 0, 1, 2, ..., 9997, 9998, 9999])
FF_decorrelated[0]_dim_0
(FF_decorrelated[0]_dim_0)
int64
0
FF_decorrelated[1]_dim_0
(FF_decorrelated[1]_dim_0)
int64
0
FF_decorrelated[2]_dim_0
(FF_decorrelated[2]_dim_0)
int64
0
FF_decorrelated[3]_dim_0
(FF_decorrelated[3]_dim_0)
int64
0
FF_decorrelated[4]_dim_0
(FF_decorrelated[4]_dim_0)
int64
0
mu_dim_0
(mu_dim_0)
int64
0
stat_dim_0
(stat_dim_0)
int64
0 1 2 3 4 5 6
array([0, 1, 2, 3, 4, 5, 6])
cvl_dim_0
(cvl_dim_0)
int64
0
csl_dim_0
(csl_dim_0)
int64
0
ct_dim_0
(ct_dim_0)
int64
0
FF_decorrelated[0]
(chain, draw, FF_decorrelated[0]_dim_0)
float64
2.015 -0.1438 ... 1.06 -0.6052
array([[[ 2.01485046],\n",
+ " [-0.14380727],\n",
+ " [-0.40579956],\n",
+ " ...,\n",
+ " [ 1.61819106],\n",
+ " [ 1.61819106],\n",
+ " [ 1.61819106]],\n",
+ "\n",
+ " [[ 0.58638077],\n",
+ " [ 0.58638077],\n",
+ " [ 0.58638077],\n",
+ " ...,\n",
+ " [ 0.13089646],\n",
+ " [ 0.13089646],\n",
+ " [ 0.28080911]],\n",
+ "\n",
+ " [[ 0.70269039],\n",
+ " [ 0.71410742],\n",
+ " [-1.06305562],\n",
+ " ...,\n",
+ " [-0.17268716],\n",
+ " [ 0.39649363],\n",
+ " [ 0.39649363]],\n",
+ "\n",
+ " [[ 0.02192634],\n",
+ " [-0.07112614],\n",
+ " [-0.07112614],\n",
+ " ...,\n",
+ " [ 0.87531679],\n",
+ " [ 1.05966546],\n",
+ " [-0.60520167]]])
FF_decorrelated[1]
(chain, draw, FF_decorrelated[1]_dim_0)
float64
0.5066 0.3234 ... 0.2023 0.2023
array([[[ 0.50661143],\n",
+ " [ 0.32338032],\n",
+ " [ 0.32338032],\n",
+ " ...,\n",
+ " [ 0.24485386],\n",
+ " [ 0.13394113],\n",
+ " [ 0.25074155]],\n",
+ "\n",
+ " [[-0.21901541],\n",
+ " [-0.70950152],\n",
+ " [-0.70950152],\n",
+ " ...,\n",
+ " [-0.69089409],\n",
+ " [-0.69089409],\n",
+ " [-0.69089409]],\n",
+ "\n",
+ " [[-1.14492442],\n",
+ " [-0.65251495],\n",
+ " [-1.8975692 ],\n",
+ " ...,\n",
+ " [-0.54605994],\n",
+ " [-0.54605994],\n",
+ " [-0.07845754]],\n",
+ "\n",
+ " [[ 0.41260391],\n",
+ " [ 0.41260391],\n",
+ " [ 0.41260391],\n",
+ " ...,\n",
+ " [ 0.20233958],\n",
+ " [ 0.20233958],\n",
+ " [ 0.20233958]]])
FF_decorrelated[2]
(chain, draw, FF_decorrelated[2]_dim_0)
float64
-1.404 -1.588 ... 0.07576 -0.6516
array([[[-1.4042462 ],\n",
+ " [-1.58776882],\n",
+ " [-1.58776882],\n",
+ " ...,\n",
+ " [-0.40461899],\n",
+ " [-1.10188132],\n",
+ " [ 0.5167337 ]],\n",
+ "\n",
+ " [[ 0.26841254],\n",
+ " [ 0.26841254],\n",
+ " [ 0.26841254],\n",
+ " ...,\n",
+ " [ 0.00183606],\n",
+ " [ 0.00183606],\n",
+ " [ 0.00183606]],\n",
+ "\n",
+ " [[-1.56427461],\n",
+ " [-0.54895782],\n",
+ " [-0.54895782],\n",
+ " ...,\n",
+ " [ 0.85612842],\n",
+ " [ 0.85612842],\n",
+ " [ 0.85612842]],\n",
+ "\n",
+ " [[-0.82637618],\n",
+ " [-1.07655602],\n",
+ " [-1.31243744],\n",
+ " ...,\n",
+ " [ 0.07575767],\n",
+ " [ 0.07575767],\n",
+ " [-0.65163336]]])
FF_decorrelated[3]
(chain, draw, FF_decorrelated[3]_dim_0)
float64
0.8381 0.8381 ... -0.7396 -0.7396
array([[[ 0.83812492],\n",
+ " [ 0.83812492],\n",
+ " [ 0.83812492],\n",
+ " ...,\n",
+ " [ 1.58853383],\n",
+ " [-0.78599873],\n",
+ " [-0.78599873]],\n",
+ "\n",
+ " [[ 0.78711278],\n",
+ " [ 0.13533011],\n",
+ " [ 0.13533011],\n",
+ " ...,\n",
+ " [ 1.0460351 ],\n",
+ " [ 1.0460351 ],\n",
+ " [ 1.0460351 ]],\n",
+ "\n",
+ " [[ 0.68223413],\n",
+ " [ 0.68223413],\n",
+ " [ 0.68223413],\n",
+ " ...,\n",
+ " [ 0.23851039],\n",
+ " [ 0.23851039],\n",
+ " [-0.60394308]],\n",
+ "\n",
+ " [[-1.54187974],\n",
+ " [-1.27245343],\n",
+ " [ 0.18949121],\n",
+ " ...,\n",
+ " [-0.73962284],\n",
+ " [-0.73962284],\n",
+ " [-0.73962284]]])
FF_decorrelated[4]
(chain, draw, FF_decorrelated[4]_dim_0)
float64
0.5649 0.3178 ... 0.2664 -1.232
array([[[ 0.56490427],\n",
+ " [ 0.31783773],\n",
+ " [ 0.31783773],\n",
+ " ...,\n",
+ " [ 1.99878998],\n",
+ " [ 2.15598103],\n",
+ " [-0.73906112]],\n",
+ "\n",
+ " [[ 0.16718579],\n",
+ " [ 0.16718579],\n",
+ " [ 0.16718579],\n",
+ " ...,\n",
+ " [-0.31891123],\n",
+ " [-1.39270624],\n",
+ " [-0.68244484]],\n",
+ "\n",
+ " [[-0.3862634 ],\n",
+ " [-0.82583209],\n",
+ " [-0.82583209],\n",
+ " ...,\n",
+ " [ 1.339876 ],\n",
+ " [-1.59521525],\n",
+ " [ 0.2481868 ]],\n",
+ "\n",
+ " [[ 0.53952994],\n",
+ " [ 1.25818407],\n",
+ " [ 1.25818407],\n",
+ " ...,\n",
+ " [ 0.26641398],\n",
+ " [ 0.26641398],\n",
+ " [-1.23193724]]])
mu
(chain, draw, mu_dim_0)
float64
1.0 1.0 1.0 1.0 ... 1.0 1.0 1.0 1.0
array([[[1.],\n",
" [1.],\n",
" [1.],\n",
" ...,\n",
@@ -659,140 +654,140 @@
" ...,\n",
" [1.],\n",
" [1.],\n",
- " [1.]]])
stat
(chain, draw, stat_dim_0)
float64
1.109 1.047 0.9959 ... 1.011 1.089
array([[[1.10923175, 1.04722043, 0.9958835 , ..., 0.93405397,\n",
- " 1.00693795, 1.09309179],\n",
- " [1.15046225, 1.04722043, 1.01768347, ..., 0.93405397,\n",
- " 1.00693795, 1.09309179],\n",
- " [1.15046225, 1.03326846, 1.01768347, ..., 0.93405397,\n",
- " 1.00693795, 1.09309179],\n",
- " ...,\n",
- " [0.97425578, 1.03847464, 1.0081129 , ..., 0.93903291,\n",
- " 1.01572505, 1.09747915],\n",
- " [0.97425578, 1.03847464, 1.02329319, ..., 0.93903291,\n",
- " 1.01572505, 1.09747915],\n",
- " [0.97425578, 1.03847464, 1.02329319, ..., 0.93835399,\n",
- " 1.01572505, 1.09747915]],\n",
- "\n",
- " [[1.04690751, 1.05955877, 1.02677238, ..., 0.93713268,\n",
- " 1.01851401, 1.09901047],\n",
- " [1.04690751, 1.05955877, 1.02677238, ..., 0.93713268,\n",
- " 1.01851401, 1.09901047],\n",
- " [1.06336905, 1.05955877, 1.02283553, ..., 0.93597627,\n",
- " 1.01851401, 1.09901047],\n",
+ " [1.]]])
stat
(chain, draw, stat_dim_0)
float64
1.026 0.9994 1.002 ... 0.9795 1.01
array([[[1.02618568, 0.99940585, 1.00226546, ..., 0.98158448,\n",
+ " 1.00270173, 1.03196902],\n",
+ " [1.02618568, 0.99036403, 1.00226546, ..., 0.9920868 ,\n",
+ " 1.00270173, 1.03196902],\n",
+ " [1.02618568, 0.99036403, 1.00226546, ..., 0.9920868 ,\n",
+ " 1.00270173, 1.03196902],\n",
+ " ...,\n",
+ " [1.01806631, 1.00381663, 1.01075693, ..., 0.9998566 ,\n",
+ " 1.00606429, 1.05948068],\n",
+ " [1.01806631, 1.00381663, 1.01075693, ..., 0.9998566 ,\n",
+ " 1.01567657, 1.05948068],\n",
+ " [1.01806631, 1.00381663, 1.01075693, ..., 0.9998566 ,\n",
+ " 1.01567657, 1.09639796]],\n",
+ "\n",
+ " [[1.0055468 , 1.00057664, 0.99256595, ..., 0.99749661,\n",
+ " 0.98794005, 1.07703231],\n",
+ " [1.0055468 , 1.00057664, 0.99256595, ..., 0.99749661,\n",
+ " 0.98794005, 1.07703231],\n",
+ " [1.0055468 , 1.00057664, 0.99256595, ..., 0.99749661,\n",
+ " 0.98794005, 1.07703231],\n",
"...\n",
- " [0.85506537, 1.04213043, 1.02187697, ..., 0.93962106,\n",
- " 1.00732042, 1.09809286],\n",
- " [0.85506537, 1.03003885, 1.02915792, ..., 0.93962106,\n",
- " 1.00732042, 1.09809286],\n",
- " [0.85506537, 1.03003885, 1.01754226, ..., 0.93962106,\n",
- " 1.00732042, 1.08807003]],\n",
- "\n",
- " [[1.04799217, 1.0344733 , 1.01929821, ..., 0.94186851,\n",
- " 1.01767742, 1.09457671],\n",
- " [1.04799217, 1.0344733 , 1.01929821, ..., 0.94186851,\n",
- " 1.01767742, 1.09457671],\n",
- " [1.04799217, 1.05894531, 1.01929821, ..., 0.94186851,\n",
- " 1.00848936, 1.09457671],\n",
- " ...,\n",
- " [0.96335814, 1.04859969, 1.00928656, ..., 0.9322577 ,\n",
- " 1.01066636, 1.08890574],\n",
- " [0.96335814, 1.03523302, 1.00928656, ..., 0.9322577 ,\n",
- " 1.01066636, 1.08890574],\n",
- " [1.06328606, 1.03523302, 1.00928656, ..., 0.9322577 ,\n",
- " 1.01066636, 1.08890574]]])
cvl
(chain, draw, cvl_dim_0)
float64
0.2203 0.2267 ... 0.2313 0.2313
array([[[0.2203403 ],\n",
- " [0.22665049],\n",
- " [0.22665049],\n",
- " ...,\n",
- " [0.24480561],\n",
- " [0.24480561],\n",
- " [0.24480561]],\n",
- "\n",
- " [[0.23505351],\n",
- " [0.23043483],\n",
- " [0.23043483],\n",
- " ...,\n",
- " [0.23840869],\n",
- " [0.23840869],\n",
- " [0.23840869]],\n",
- "\n",
- " [[0.18300133],\n",
- " [0.18300133],\n",
- " [0.18300133],\n",
- " ...,\n",
- " [0.22494899],\n",
- " [0.22494899],\n",
- " [0.22494899]],\n",
- "\n",
- " [[0.22441878],\n",
- " [0.22441878],\n",
- " [0.22441878],\n",
- " ...,\n",
- " [0.23129876],\n",
- " [0.23129876],\n",
- " [0.23129876]]])
csl
(chain, draw, csl_dim_0)
float64
0.779 0.779 0.779 ... 0.7782 0.7782
array([[[0.77897849],\n",
- " [0.77897849],\n",
- " [0.77897849],\n",
- " ...,\n",
- " [0.77315922],\n",
- " [0.77315922],\n",
- " [0.77145696]],\n",
- "\n",
- " [[0.77335727],\n",
- " [0.77335727],\n",
- " [0.77335727],\n",
- " ...,\n",
- " [0.7702336 ],\n",
- " [0.7702336 ],\n",
- " [0.7702336 ]],\n",
- "\n",
- " [[0.80272633],\n",
- " [0.80272633],\n",
- " [0.80272633],\n",
- " ...,\n",
- " [0.77551396],\n",
- " [0.77551396],\n",
- " [0.77551396]],\n",
- "\n",
- " [[0.77278283],\n",
- " [0.77278283],\n",
- " [0.77278283],\n",
- " ...,\n",
- " [0.778809 ],\n",
- " [0.77815407],\n",
- " [0.77815407]]])
ct
(chain, draw, ct_dim_0)
float64
0.03225 0.03225 ... 0.01907 0.01907
array([[[0.03225057],\n",
- " [0.03225057],\n",
- " [0.03801888],\n",
- " ...,\n",
- " [0.00944647],\n",
- " [0.00984188],\n",
- " [0.00977739]],\n",
- "\n",
- " [[0.02537348],\n",
- " [0.02537348],\n",
- " [0.02537348],\n",
- " ...,\n",
- " [0.01802346],\n",
- " [0.02475649],\n",
- " [0.01076715]],\n",
- "\n",
- " [[0.05020308],\n",
- " [0.05337035],\n",
- " [0.04914154],\n",
- " ...,\n",
- " [0.03642157],\n",
- " [0.0349287 ],\n",
- " [0.0349287 ]],\n",
- "\n",
- " [[0.03809965],\n",
- " [0.03809965],\n",
- " [0.03809965],\n",
- " ...,\n",
- " [0.02956821],\n",
- " [0.01907469],\n",
- " [0.01907469]]])
PandasIndex
PandasIndex(Index([0, 1, 2, 3], dtype='int64', name='chain'))
PandasIndex
PandasIndex(Index([ 0, 1, 2, 3, 4, 5, 6, 7, 8, 9,\n",
+ " [0.99182728, 1.00588749, 1.00194187, ..., 1.02075054,\n",
+ " 1.00526187, 1.13104207],\n",
+ " [0.99182728, 1.00588749, 1.00674914, ..., 1.01205988,\n",
+ " 1.00526187, 1.13104207],\n",
+ " [0.99182728, 1.00588749, 1.00674914, ..., 1.01229097,\n",
+ " 0.97992747, 1.13104207]],\n",
+ "\n",
+ " [[1.00704466, 0.9955772 , 1.00428955, ..., 0.98691818,\n",
+ " 0.95850821, 1.01113249],\n",
+ " [1.00704466, 0.9955772 , 1.00428955, ..., 0.98691818,\n",
+ " 0.95850821, 1.01113249],\n",
+ " [1.00704466, 0.9955772 , 1.00428955, ..., 1.0062064 ,\n",
+ " 0.95038563, 1.01113249],\n",
+ " ...,\n",
+ " [1.02657296, 1.00362554, 1.00106032, ..., 1.01075644,\n",
+ " 0.97948684, 1.01049664],\n",
+ " [0.97558766, 1.00362554, 1.00106032, ..., 1.01075644,\n",
+ " 0.97948684, 1.01049664],\n",
+ " [0.97558766, 1.00362554, 1.00106032, ..., 1.01075644,\n",
+ " 0.97948684, 1.01049664]]])
cvl
(chain, draw, cvl_dim_0)
float64
0.7268 0.7268 ... 0.9691 0.9691
array([[[0.72680633],\n",
+ " [0.72680633],\n",
+ " [0.72680633],\n",
+ " ...,\n",
+ " [0.81107748],\n",
+ " [0.81107748],\n",
+ " [0.81412342]],\n",
+ "\n",
+ " [[0.86507951],\n",
+ " [0.8693156 ],\n",
+ " [0.8693156 ],\n",
+ " ...,\n",
+ " [0.95707061],\n",
+ " [0.95707061],\n",
+ " [0.95707061]],\n",
+ "\n",
+ " [[0.57226442],\n",
+ " [0.57226442],\n",
+ " [0.57226442],\n",
+ " ...,\n",
+ " [1.00993693],\n",
+ " [1.00993693],\n",
+ " [1.00993693]],\n",
+ "\n",
+ " [[0.89123486],\n",
+ " [0.89123486],\n",
+ " [0.89123486],\n",
+ " ...,\n",
+ " [0.96913649],\n",
+ " [0.96913649],\n",
+ " [0.96913649]]])
csl
(chain, draw, csl_dim_0)
float64
0.508 0.508 0.508 ... 0.5301 0.5301
array([[[0.50798671],\n",
+ " [0.50798671],\n",
+ " [0.50798671],\n",
+ " ...,\n",
+ " [0.49673649],\n",
+ " [0.49673649],\n",
+ " [0.49673649]],\n",
+ "\n",
+ " [[0.50842695],\n",
+ " [0.51076928],\n",
+ " [0.51076928],\n",
+ " ...,\n",
+ " [0.51296871],\n",
+ " [0.51296871],\n",
+ " [0.51296871]],\n",
+ "\n",
+ " [[0.49509846],\n",
+ " [0.49509846],\n",
+ " [0.49509846],\n",
+ " ...,\n",
+ " [0.48283193],\n",
+ " [0.48283193],\n",
+ " [0.48283193]],\n",
+ "\n",
+ " [[0.55291641],\n",
+ " [0.55483858],\n",
+ " [0.55057329],\n",
+ " ...,\n",
+ " [0.53008771],\n",
+ " [0.53008771],\n",
+ " [0.53008771]]])
ct
(chain, draw, ct_dim_0)
float64
0.7451 0.7419 ... 0.498 0.498
array([[[0.74505913],\n",
+ " [0.74190119],\n",
+ " [0.74190119],\n",
+ " ...,\n",
+ " [0.66167168],\n",
+ " [0.66167168],\n",
+ " [0.66167168]],\n",
+ "\n",
+ " [[0.61821406],\n",
+ " [0.61821406],\n",
+ " [0.60898689],\n",
+ " ...,\n",
+ " [0.52209135],\n",
+ " [0.52209135],\n",
+ " [0.52209135]],\n",
+ "\n",
+ " [[0.87949373],\n",
+ " [0.88135554],\n",
+ " [0.88135554],\n",
+ " ...,\n",
+ " [0.48479212],\n",
+ " [0.48845117],\n",
+ " [0.48845117]],\n",
+ "\n",
+ " [[0.56909604],\n",
+ " [0.56909604],\n",
+ " [0.56909604],\n",
+ " ...,\n",
+ " [0.49800785],\n",
+ " [0.49800785],\n",
+ " [0.49800785]]])
PandasIndex
PandasIndex(Index([0, 1, 2, 3], dtype='int64', name='chain'))
PandasIndex
PandasIndex(Index([ 0, 1, 2, 3, 4, 5, 6, 7, 8, 9,\n",
" ...\n",
" 9990, 9991, 9992, 9993, 9994, 9995, 9996, 9997, 9998, 9999],\n",
- " dtype='int64', name='draw', length=10000))
PandasIndex
PandasIndex(Index([0], dtype='int64', name='FF_decorrelated[0]_dim_0'))
PandasIndex
PandasIndex(Index([0], dtype='int64', name='FF_decorrelated[1]_dim_0'))
PandasIndex
PandasIndex(Index([0], dtype='int64', name='FF_decorrelated[2]_dim_0'))
PandasIndex
PandasIndex(Index([0], dtype='int64', name='FF_decorrelated[3]_dim_0'))
PandasIndex
PandasIndex(Index([0], dtype='int64', name='FF_decorrelated[4]_dim_0'))
PandasIndex
PandasIndex(Index([0], dtype='int64', name='mu_dim_0'))
PandasIndex
PandasIndex(Index([0, 1, 2, 3, 4, 5, 6], dtype='int64', name='stat_dim_0'))
PandasIndex
PandasIndex(Index([0], dtype='int64', name='cvl_dim_0'))
PandasIndex
PandasIndex(Index([0], dtype='int64', name='csl_dim_0'))
PandasIndex
PandasIndex(Index([0], dtype='int64', name='ct_dim_0'))
- created_at :
- 2023-11-24T19:48:21.837970
- arviz_version :
- 0.16.1
- inference_library :
- pymc
- inference_library_version :
- 5.9.2
- sampling_time :
- 8654.44973897934
- tuning_steps :
- 1500
"
+ " dtype='int64', name='draw', length=10000))
PandasIndex
PandasIndex(Index([0], dtype='int64', name='FF_decorrelated[0]_dim_0'))
PandasIndex
PandasIndex(Index([0], dtype='int64', name='FF_decorrelated[1]_dim_0'))
PandasIndex
PandasIndex(Index([0], dtype='int64', name='FF_decorrelated[2]_dim_0'))
PandasIndex
PandasIndex(Index([0], dtype='int64', name='FF_decorrelated[3]_dim_0'))
PandasIndex
PandasIndex(Index([0], dtype='int64', name='FF_decorrelated[4]_dim_0'))
PandasIndex
PandasIndex(Index([0], dtype='int64', name='mu_dim_0'))
PandasIndex
PandasIndex(Index([0, 1, 2, 3, 4, 5, 6], dtype='int64', name='stat_dim_0'))
PandasIndex
PandasIndex(Index([0], dtype='int64', name='cvl_dim_0'))
PandasIndex
PandasIndex(Index([0], dtype='int64', name='csl_dim_0'))
PandasIndex
PandasIndex(Index([0], dtype='int64', name='ct_dim_0'))
- created_at :
- 2023-11-27T17:35:25.353355
- arviz_version :
- 0.16.1
- inference_library :
- pymc
- inference_library_version :
- 5.10.0
- sampling_time :
- 6302.489182949066
- tuning_steps :
- 1500
"
],
"text/plain": [
"\n",
@@ -824,16 +819,16 @@
" FF_decorrelated[3] (chain, draw, FF_decorrelated[3]_dim_0) float64 ...\n",
" FF_decorrelated[4] (chain, draw, FF_decorrelated[4]_dim_0) float64 ...\n",
" mu (chain, draw, mu_dim_0) float64 1.0 1.0 ... 1.0\n",
- " stat (chain, draw, stat_dim_0) float64 1.109 ... 1.089\n",
- " cvl (chain, draw, cvl_dim_0) float64 0.2203 ... 0.2313\n",
- " csl (chain, draw, csl_dim_0) float64 0.779 ... 0.7782\n",
- " ct (chain, draw, ct_dim_0) float64 0.03225 ... 0.0...\n",
+ " stat (chain, draw, stat_dim_0) float64 1.026 ... 1.01\n",
+ " cvl (chain, draw, cvl_dim_0) float64 0.7268 ... 0.9691\n",
+ " csl (chain, draw, csl_dim_0) float64 0.508 ... 0.5301\n",
+ " ct (chain, draw, ct_dim_0) float64 0.7451 ... 0.498\n",
"Attributes:\n",
- " created_at: 2023-11-24T19:48:21.837970\n",
+ " created_at: 2023-11-27T17:35:25.353355\n",
" arviz_version: 0.16.1\n",
" inference_library: pymc\n",
- " inference_library_version: 5.9.2\n",
- " sampling_time: 8654.44973897934\n",
+ " inference_library_version: 5.10.0\n",
+ " sampling_time: 6302.489182949066\n",
" tuning_steps: 1500"
]
},
@@ -848,12 +843,12 @@
},
{
"cell_type": "code",
- "execution_count": 6,
+ "execution_count": 7,
"metadata": {},
"outputs": [
{
"data": {
- "image/png": "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",
+ "image/png": "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",
"text/plain": [
"