Skip to content

Commit 941ccb5

Browse files
bartzbeielsteinbartzbeielstein
committed
v0.9.7
1 parent dcee62e commit 941ccb5

5 files changed

Lines changed: 260 additions & 14 deletions

File tree

‎notebooks/00_spotPython_tests.ipynb‎

Lines changed: 209 additions & 4 deletions
Original file line numberDiff line numberDiff line change
@@ -1527,7 +1527,7 @@
15271527
},
15281528
{
15291529
"cell_type": "code",
1530-
"execution_count": 16,
1530+
"execution_count": 1,
15311531
"metadata": {},
15321532
"outputs": [
15331533
{
@@ -1680,12 +1680,217 @@
16801680
" break"
16811681
]
16821682
},
1683+
{
1684+
"cell_type": "markdown",
1685+
"metadata": {},
1686+
"source": [
1687+
"## Test HyperLight"
1688+
]
1689+
},
16831690
{
16841691
"cell_type": "code",
1685-
"execution_count": null,
1692+
"execution_count": 3,
16861693
"metadata": {},
1687-
"outputs": [],
1688-
"source": []
1694+
"outputs": [
1695+
{
1696+
"name": "stderr",
1697+
"output_type": "stream",
1698+
"text": [
1699+
"Seed set to 1234\n"
1700+
]
1701+
},
1702+
{
1703+
"data": {
1704+
"text/plain": [
1705+
"array([[ True, True, True, True, True, True, True, True, True],\n",
1706+
" [ True, True, True, True, True, True, True, True, True]])"
1707+
]
1708+
},
1709+
"execution_count": 3,
1710+
"metadata": {},
1711+
"output_type": "execute_result"
1712+
}
1713+
],
1714+
"source": [
1715+
"import numpy as np\n",
1716+
"from spotPython.utils.init import fun_control_init\n",
1717+
"from spotPython.light.regression.netlightregression import NetLightRegression\n",
1718+
"from spotPython.hyperdict.light_hyper_dict import LightHyperDict\n",
1719+
"from spotPython.hyperparameters.values import add_core_model_to_fun_control\n",
1720+
"from spotPython.fun.hyperlight import HyperLight\n",
1721+
"from spotPython.hyperparameters.values import get_var_name\n",
1722+
"fun_control = fun_control_init()\n",
1723+
"add_core_model_to_fun_control(core_model=NetLightRegression,\n",
1724+
" fun_control=fun_control,\n",
1725+
" hyper_dict=LightHyperDict)\n",
1726+
"hyper_light = HyperLight(seed=126, log_level=50)\n",
1727+
"n_hyperparams = len(get_var_name(fun_control))\n",
1728+
"# generate a random np.array X with shape (2, n_hyperparams)\n",
1729+
"X = np.random.rand(2, n_hyperparams)\n",
1730+
"X == hyper_light.check_X_shape(X, fun_control)"
1731+
]
1732+
},
1733+
{
1734+
"cell_type": "code",
1735+
"execution_count": 7,
1736+
"metadata": {},
1737+
"outputs": [
1738+
{
1739+
"name": "stderr",
1740+
"output_type": "stream",
1741+
"text": [
1742+
"Seed set to 1234\n",
1743+
"/Users/bartz/miniforge3/envs/spotCondaEnv/lib/python3.11/site-packages/lightning/pytorch/utilities/parsing.py:198: Attribute 'act_fn' is an instance of `nn.Module` and is already saved during checkpointing. It is recommended to ignore them using `self.save_hyperparameters(ignore=['act_fn'])`.\n",
1744+
"GPU available: True (mps), used: True\n",
1745+
"TPU available: False, using: 0 TPU cores\n",
1746+
"IPU available: False, using: 0 IPUs\n",
1747+
"HPU available: False, using: 0 HPUs\n",
1748+
"/Users/bartz/miniforge3/envs/spotCondaEnv/lib/python3.11/site-packages/lightning/pytorch/callbacks/model_checkpoint.py:639: Checkpoint directory runs/lightning_logs/-2621019871347415348/checkpoints exists and is not empty.\n",
1749+
"\n",
1750+
" | Name | Type | Params | In sizes | Out sizes\n",
1751+
"-------------------------------------------------------------\n",
1752+
"0 | layers | Sequential | 157 | [16, 10] | [16, 1] \n",
1753+
"-------------------------------------------------------------\n",
1754+
"157 Trainable params\n",
1755+
"0 Non-trainable params\n",
1756+
"157 Total params\n",
1757+
"0.001 Total estimated model params size (MB)\n",
1758+
"/Users/bartz/miniforge3/envs/spotCondaEnv/lib/python3.11/site-packages/lightning/pytorch/trainer/connectors/data_connector.py:441: The 'val_dataloader' does not have many workers which may be a bottleneck. Consider increasing the value of the `num_workers` argument` to `num_workers=11` in the `DataLoader` to improve performance.\n",
1759+
"/Users/bartz/miniforge3/envs/spotCondaEnv/lib/python3.11/site-packages/lightning/pytorch/trainer/connectors/data_connector.py:441: The 'train_dataloader' does not have many workers which may be a bottleneck. Consider increasing the value of the `num_workers` argument` to `num_workers=11` in the `DataLoader` to improve performance.\n",
1760+
"/Users/bartz/miniforge3/envs/spotCondaEnv/lib/python3.11/site-packages/lightning/pytorch/loops/fit_loop.py:293: The number of training batches (5) is smaller than the logging interval Trainer(log_every_n_steps=50). Set a lower value for log_every_n_steps if you want to see logs for the training epoch.\n"
1761+
]
1762+
},
1763+
{
1764+
"name": "stdout",
1765+
"output_type": "stream",
1766+
"text": [
1767+
"Train_model(): Test set size: 266\n"
1768+
]
1769+
},
1770+
{
1771+
"data": {
1772+
"text/html": [
1773+
"<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\">┏━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━┓\n",
1774+
"┃<span style=\"font-weight: bold\"> Validate metric </span>┃<span style=\"font-weight: bold\"> DataLoader 0 </span>┃\n",
1775+
"┡━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━┩\n",
1776+
"│<span style=\"color: #008080; text-decoration-color: #008080\"> hp_metric </span>│<span style=\"color: #800080; text-decoration-color: #800080\"> 27259.2421875 </span>│\n",
1777+
"│<span style=\"color: #008080; text-decoration-color: #008080\"> val_loss </span>│<span style=\"color: #800080; text-decoration-color: #800080\"> 27259.2421875 </span>│\n",
1778+
"└───────────────────────────┴───────────────────────────┘\n",
1779+
"</pre>\n"
1780+
],
1781+
"text/plain": [
1782+
"┏━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━┓\n",
1783+
"┃\u001b[1m \u001b[0m\u001b[1m Validate metric \u001b[0m\u001b[1m \u001b[0m┃\u001b[1m \u001b[0m\u001b[1m DataLoader 0 \u001b[0m\u001b[1m \u001b[0m┃\n",
1784+
"┡━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━┩\n",
1785+
"│\u001b[36m \u001b[0m\u001b[36m hp_metric \u001b[0m\u001b[36m \u001b[0m│\u001b[35m \u001b[0m\u001b[35m 27259.2421875 \u001b[0m\u001b[35m \u001b[0m│\n",
1786+
"│\u001b[36m \u001b[0m\u001b[36m val_loss \u001b[0m\u001b[36m \u001b[0m│\u001b[35m \u001b[0m\u001b[35m 27259.2421875 \u001b[0m\u001b[35m \u001b[0m│\n",
1787+
"└───────────────────────────┴───────────────────────────┘\n"
1788+
]
1789+
},
1790+
"metadata": {},
1791+
"output_type": "display_data"
1792+
},
1793+
{
1794+
"name": "stderr",
1795+
"output_type": "stream",
1796+
"text": [
1797+
"GPU available: True (mps), used: True\n",
1798+
"TPU available: False, using: 0 TPU cores\n",
1799+
"IPU available: False, using: 0 IPUs\n",
1800+
"HPU available: False, using: 0 HPUs\n",
1801+
"\n",
1802+
" | Name | Type | Params | In sizes | Out sizes\n",
1803+
"-------------------------------------------------------------\n",
1804+
"0 | layers | Sequential | 157 | [16, 10] | [16, 1] \n",
1805+
"-------------------------------------------------------------\n",
1806+
"157 Trainable params\n",
1807+
"0 Non-trainable params\n",
1808+
"157 Total params\n",
1809+
"0.001 Total estimated model params size (MB)\n"
1810+
]
1811+
},
1812+
{
1813+
"name": "stdout",
1814+
"output_type": "stream",
1815+
"text": [
1816+
"train_model result: {'val_loss': 27259.2421875, 'hp_metric': 27259.2421875}\n",
1817+
"Train_model(): Test set size: 266\n"
1818+
]
1819+
},
1820+
{
1821+
"name": "stderr",
1822+
"output_type": "stream",
1823+
"text": [
1824+
"`Trainer.fit` stopped: `max_epochs=16` reached.\n"
1825+
]
1826+
},
1827+
{
1828+
"data": {
1829+
"text/html": [
1830+
"<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\">┏━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━┓\n",
1831+
"┃<span style=\"font-weight: bold\"> Validate metric </span>┃<span style=\"font-weight: bold\"> DataLoader 0 </span>┃\n",
1832+
"┡━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━┩\n",
1833+
"│<span style=\"color: #008080; text-decoration-color: #008080\"> hp_metric </span>│<span style=\"color: #800080; text-decoration-color: #800080\"> 24221.275390625 </span>│\n",
1834+
"│<span style=\"color: #008080; text-decoration-color: #008080\"> val_loss </span>│<span style=\"color: #800080; text-decoration-color: #800080\"> 24221.275390625 </span>│\n",
1835+
"└───────────────────────────┴───────────────────────────┘\n",
1836+
"</pre>\n"
1837+
],
1838+
"text/plain": [
1839+
"┏━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━┓\n",
1840+
"┃\u001b[1m \u001b[0m\u001b[1m Validate metric \u001b[0m\u001b[1m \u001b[0m┃\u001b[1m \u001b[0m\u001b[1m DataLoader 0 \u001b[0m\u001b[1m \u001b[0m┃\n",
1841+
"┡━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━┩\n",
1842+
"│\u001b[36m \u001b[0m\u001b[36m hp_metric \u001b[0m\u001b[36m \u001b[0m│\u001b[35m \u001b[0m\u001b[35m 24221.275390625 \u001b[0m\u001b[35m \u001b[0m│\n",
1843+
"│\u001b[36m \u001b[0m\u001b[36m val_loss \u001b[0m\u001b[36m \u001b[0m│\u001b[35m \u001b[0m\u001b[35m 24221.275390625 \u001b[0m\u001b[35m \u001b[0m│\n",
1844+
"└───────────────────────────┴───────────────────────────┘\n"
1845+
]
1846+
},
1847+
"metadata": {},
1848+
"output_type": "display_data"
1849+
},
1850+
{
1851+
"name": "stdout",
1852+
"output_type": "stream",
1853+
"text": [
1854+
"train_model result: {'val_loss': 24221.275390625, 'hp_metric': 24221.275390625}\n"
1855+
]
1856+
},
1857+
{
1858+
"data": {
1859+
"text/plain": [
1860+
"array([27259.2421875 , 24221.27539062])"
1861+
]
1862+
},
1863+
"execution_count": 7,
1864+
"metadata": {},
1865+
"output_type": "execute_result"
1866+
}
1867+
],
1868+
"source": [
1869+
"from spotPython.utils.init import fun_control_init\n",
1870+
"from spotPython.light.regression.netlightregression import NetLightRegression\n",
1871+
"from spotPython.hyperdict.light_hyper_dict import LightHyperDict\n",
1872+
"from spotPython.hyperparameters.values import (add_core_model_to_fun_control,\n",
1873+
" get_default_hyperparameters_as_array)\n",
1874+
"from spotPython.fun.hyperlight import HyperLight\n",
1875+
"from spotPython.data.diabetes import Diabetes\n",
1876+
"from spotPython.hyperparameters.values import set_data_set\n",
1877+
"import numpy as np\n",
1878+
"fun_control = fun_control_init(\n",
1879+
" _L_in=10,\n",
1880+
" _L_out=1,)\n",
1881+
"dataset = Diabetes()\n",
1882+
"set_data_set(fun_control=fun_control,\n",
1883+
" data_set=dataset)\n",
1884+
"add_core_model_to_fun_control(core_model=NetLightRegression,\n",
1885+
" fun_control=fun_control,\n",
1886+
" hyper_dict=LightHyperDict)\n",
1887+
"hyper_light = HyperLight(seed=126, log_level=50)\n",
1888+
"X = get_default_hyperparameters_as_array(fun_control)\n",
1889+
"# combine X and X to a np.array with shape (2, n_hyperparams)\n",
1890+
"# so that two values are returned\n",
1891+
"X = np.vstack((X, X))\n",
1892+
"hyper_light.fun(X, fun_control)\n"
1893+
]
16891894
},
16901895
{
16911896
"cell_type": "code",

‎pyproject.toml‎

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -7,7 +7,7 @@ build-backend = "setuptools.build_meta"
77

88
[project]
99
name = "spotPython"
10-
version = "0.9.6"
10+
version = "0.9.7"
1111
authors = [
1212
{ name="T. Bartz-Beielstein", email="tbb@bartzundbartz.de" }
1313
]

‎src/spotPython/build/kriging.py‎

Lines changed: 36 additions & 3 deletions
Original file line numberDiff line numberDiff line change
@@ -476,6 +476,8 @@ def fit(self, nat_X: np.ndarray, nat_y: np.ndarray) -> object:
476476
[1. 1.00000001]]
477477
478478
"""
479+
logger.debug("In fit(): nat_X: %s", nat_X)
480+
logger.debug("In fit(): nat_y: %s", nat_y)
479481
self.initialize_variables(nat_X, nat_y)
480482
self.set_variable_types()
481483
self.set_theta_values()
@@ -539,6 +541,13 @@ def initialize_variables(self, nat_X: np.ndarray, nat_y: np.ndarray) -> None:
539541
# aggregated y values:
540542
mu = Z[1]
541543
self.aggregated_mean_y = np.copy(mu)
544+
logger.debug("In initialize_variables(): self.nat_X: %s", self.nat_X)
545+
logger.debug("In initialize_variables(): self.nat_y: %s", self.nat_y)
546+
logger.debug("In initialize_variables(): self.aggregated_mean_y: %s", self.aggregated_mean_y)
547+
logger.debug("In initialize_variables(): self.min_X: %s", self.min_X)
548+
logger.debug("In initialize_variables(): self.max_X: %s", self.max_X)
549+
logger.debug("In initialize_variables(): self.n: %s", self.n)
550+
logger.debug("In initialize_variables(): self.k: %s", self.k)
542551

543552
def set_variable_types(self) -> None:
544553
"""
@@ -573,15 +582,22 @@ def set_variable_types(self) -> None:
573582
Returns:
574583
None
575584
"""
585+
logger.debug("In set_variable_types(): self.k: %s", self.k)
586+
logger.debug("In set_variable_types(): self.var_type: %s", self.var_type)
576587
# assume all variable types are "num" if "num" is
577588
# specified once:
578589
if len(self.var_type) < self.k:
579590
self.var_type = self.var_type * self.k
580-
logger.warning("Warning: All variable types forced to 'num'.")
591+
logger.warning("In set_variable_types(): All variable types forced to 'num'.")
592+
logger.debug("In set_variable_types(): self.var_type: %s", self.var_type)
581593
self.num_mask = np.array(list(map(lambda x: x == "num", self.var_type)))
582594
self.factor_mask = np.array(list(map(lambda x: x == "factor", self.var_type)))
583595
self.int_mask = np.array(list(map(lambda x: x == "int", self.var_type)))
584596
self.ordered_mask = np.array(list(map(lambda x: x == "int" or x == "num" or x == "float", self.var_type)))
597+
logger.debug("In set_variable_types(): self.num_mask: %s", self.num_mask)
598+
logger.debug("In set_variable_types(): self.factor_mask: %s", self.factor_mask)
599+
logger.debug("In set_variable_types(): self.int_mask: %s", self.int_mask)
600+
logger.debug("In set_variable_types(): self.ordered_mask: %s", self.ordered_mask)
585601

586602
def set_theta_values(self) -> None:
587603
"""
@@ -614,13 +630,19 @@ def set_theta_values(self) -> None:
614630
S.set_theta_values()
615631
assert S.theta.all() == array([0., 0.]).all()
616632
"""
617-
if self.n_theta > self.k:
633+
logger.debug("In set_theta_values(): self.k: %s", self.k)
634+
logger.debug("In set_theta_values(): self.n_theta: %s", self.n_theta)
635+
if self.n_theta > 1 or self.n_theta > self.k:
618636
self.n_theta = self.k
619-
logger.warning("More theta values than dimensions. `n_theta` set to `k`.")
637+
logger.warning("Too few theta values or more theta values than dimensions. `n_theta` set to `k`.")
638+
logger.debug("In set_theta_values(): self.n_theta: %s", self.n_theta)
620639
if self.theta_init_zero:
621640
self.theta: List[float] = zeros(self.n_theta)
641+
logger.debug("In set_theta_values(): self.theta: %s", self.theta)
622642
else:
643+
logger.debug("In set_theta_values(): self.n: %s", self.n)
623644
self.theta: List[float] = ones((self.n_theta,)) * self.n / (100 * self.k)
645+
logger.debug("In set_theta_values(): self.theta: %s", self.theta)
624646

625647
def initialize_matrices(self) -> None:
626648
"""
@@ -660,19 +682,30 @@ def initialize_matrices(self) -> None:
660682
Returns:
661683
None
662684
"""
685+
logger.debug("In initialize_matrices(): self.n_p: %s", self.n_p)
663686
self.p = ones(self.n_p) * 2.0
687+
logger.debug("In initialize_matrices(): self.p: %s", self.p)
664688
# if var(self.nat_y) is > 0, then self.pen_val = self.n * log(var(self.nat_y)) + 1e4
665689
# else self.pen_val = self.n * var(self.nat_y) + 1e4
690+
logger.debug("In initialize_matrices(): var(self.nat_y): %s", var(self.nat_y))
691+
logger.debug("In initialize_matrices(): self.n: %s", self.n)
666692
if var(self.nat_y) > 0:
667693
self.pen_val = self.n * log(var(self.nat_y)) + 1e4
668694
else:
669695
self.pen_val = self.n * var(self.nat_y) + 1e4
696+
logger.debug("In initialize_matrices(): self.pen_val: %s", self.pen_val)
670697
self.negLnLike = None
698+
logger.debug("In initialize_matrices(): self.k: %s", self.k)
699+
logger.debug("In initialize_matrices(): self.seed: %s", self.seed)
671700
self.gen = spacefilling(k=self.k, seed=self.seed)
701+
logger.debug("In initialize_matrices(): self.gen: %s", self.gen)
672702
self.LnDetPsi = None
673703
self.Psi = zeros((self.n, self.n), dtype=float64)
704+
logger.debug("In initialize_matrices(): self.Psi: %s", self.Psi)
674705
self.psi = zeros((self.n, 1))
706+
logger.debug("In initialize_matrices(): self.psi: %s", self.psi)
675707
self.one = ones(self.n)
708+
logger.debug("In initialize_matrices(): self.one: %s", self.one)
676709
self.mu = None
677710
self.U = None
678711
self.SigmaSqr = None

0 commit comments

Comments
 (0)