SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2206.10697 · 2022

Performance Prediction Under Dataset Shift

arXiv · PDF · Open in the Atlas

Code that ran

14 of 16 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (0 of them constructed an object rather than computing a result; 0 honoured the contract we drafted, 0 violated it and 14 have no contract verdict; 0 failures of our instrument; 2 no recorded run), measured 2026-11-22; repository links from the Papers with Code archive attributed under CC BY-SA, no affiliation; bodies inline only under a permissive licence as GitHub detects it. Not a reproduction of the paper's results.

"Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results. This record, dated, as JSON: /api/ran/2206.10697.json · badge: /api/ran/2206.10697.svg

Each repository's role is a label: a classifier's output recorded on the link, or, where no role is recorded, what the Papers with Code archive marked. The line under it names its basis. Neither says who wrote the code.

RepositoryRoleRan
dataiku-research/performance_prediction_under_shift canonical (classifier label)
classifier basis not recorded on the link
14 ran · 2 no recorded run
FunctionStatusWhere it lives
any_other_label Ran
ran on a synthesized input
dataiku-research/performance_prediction_under_shift/drift_dac/perturbation_shared_utils.py
code served (permissive licence) · get_code("005d05c5fb2709d2")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned int64, ndim 0, dtype family int, shape () as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import numpy as np

# Set seed for reproducibility
np.random.seed(0)

# Create a numpy array of labels
labels = np.array([0, 1, 2, 3, 4])

# Define label_to_change - should be one of the values in labels
label_to_change = 1

# Construct the ARGS dict
ARGS = {
    "label_to_change": label_to_change,
    "labels": labels,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type int64, ndim 0, dtype family int, shape () — shape () as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
get_X_y Ran
ran on a synthesized input
dataiku-research/performance_prediction_under_shift/drift_dac_experiments/perf_drop_experiment_utils.py
code served (permissive licence) · get_code("2b7902ac234fdd50")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned tuple. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import numpy as np
import pandas as pd

# Create a simple DataFrame with numeric data
data = {
    'feature1': [1.0, 2.0],
    'feature2': [3.0, 4.0],
    'feature3': [5.0, 6.0],
    'target': [7.0, 8.0]
}
df = pd.DataFrame(data)

ARGS = {
    'df': df,
    'target': 'target'
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, pandas — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type tuple — a tuple output has no array shape of its own; none is stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
knockout_shift Ran
ran on a synthesized input
dataiku-research/performance_prediction_under_shift/drift_dac/prior_shift.py
code served (permissive licence) · get_code("8dcd025f2071ff67")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned tuple. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import numpy as np

np.random.seed(0)

# Create sample data
x = np.random.randn(10, 4)  # 10 samples, 4 features
y = np.array([0, 0, 0, 0, 1, 1, 1, 2, 2, 2])  # 10 labels with class 0 appearing 4 times

# Parameters for knockout_shift
cl = 0  # class label to knockout
delta = 0.5  # fraction of samples to delete

ARGS = {
    'x': x,
    'y': y,
    'cl': cl,
    'delta': delta
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type tuple — a tuple output has no array shape of its own; none is stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
only_one_shift Ran
ran on a synthesized input
dataiku-research/performance_prediction_under_shift/drift_dac/prior_shift.py
code served (permissive licence) · get_code("d81e69bf33734e70")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned tuple. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import numpy as np

# Set random seed for reproducibility
np.random.seed(0)

# Create sample data
# x: feature matrix with shape (n_samples, n_features)
# y: label vector with shape (n_samples,)
x = np.array([
    [1.0, 2.0, 3.0, 4.0],
    [5.0, 6.0, 7.0, 8.0],
    [9.0, 10.0, 11.0, 12.0],
    [13.0, 14.0, 15.0, 16.0],
    [17.0, 18.0, 19.0, 20.0],
    [21.0, 22.0, 23.0, 24.0],
], dtype=np.float32)

# y contains multiple class labels
y = np.array([0, 1, 0, 1, 2, 2], dtype=np.int32)

# keep_cl: the class label to keep (others will be knocked out)
keep_cl = 0

ARGS = {
    "x": x,
    "y": y,
    "keep_cl": keep_cl,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type tuple — a tuple output has no array shape of its own; none is stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
prediction_drift Ran
ran on a synthesized input
dataiku-research/performance_prediction_under_shift/drift_dac/drift_metrics.py
code served (permissive licence) · get_code("ae1884f61848e0f1")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned tuple, ndim 1, dtype family float; no output shape is stated. The image it ran in was not recorded. If your copy returns the same type and ndim on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import numpy as np
import random

random.seed(0)
np.random.seed(0)

# Create sample predictions for source domain
# Shape: (batch_size, num_classes)
# batch_size=2, num_classes=4
predictions_src = np.array([
    [0.7, 0.1, 0.1, 0.1],
    [0.1, 0.8, 0.05, 0.05]
], dtype=np.float32)

# Create sample predictions for target domain
# Same shape as predictions_src
predictions_tar = np.array([
    [0.6, 0.2, 0.1, 0.1],
    [0.2, 0.7, 0.05, 0.05]
], dtype=np.float32)

ARGS = {
    "predictions_src": predictions_src,
    "predictions_tar": predictions_tar
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, scipy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type tuple, ndim 1, dtype family float — a tuple output has no array shape of its own; none is stated
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
rebalance_shift Ran
ran on a synthesized input
dataiku-research/performance_prediction_under_shift/drift_dac/prior_shift.py
code served (permissive licence) · get_code("f70c0142ed4deeba")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned tuple. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import numpy as np

np.random.seed(0)

# Create a small dataset with 2 samples and 4 features
x = np.array([
    [1.0, 2.0, 3.0, 4.0],
    [5.0, 6.0, 7.0, 8.0]
], dtype=np.float32)

# Create labels for 2 classes
y = np.array([0, 1], dtype=np.int32)

# Create priors that sum to 1.0
priors = np.array([0.6, 0.4], dtype=np.float32)

ARGS = {
    "x": x,
    "y": y,
    "priors": priors
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type tuple — a tuple output has no array shape of its own; none is stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
resize_dataset Ran
ran on a synthesized input
dataiku-research/performance_prediction_under_shift/drift_dac_experiments/multi_domain_performance_predictor.py
code served (permissive licence) · get_code("0a325b447708425d")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned tuple. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import numpy as np

# Create small test data
np.random.seed(0)
X = np.random.randn(10, 4)  # 10 samples, 4 features
y = np.random.randint(0, 2, size=10)  # 10 labels

ARGS = {
    "X": X,
    "y": y,
    "n_samples_out": 500
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type tuple — a tuple output has no array shape of its own; none is stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
retrieve_domain_df Ran
ran on a synthesized input
dataiku-research/performance_prediction_under_shift/drift_dac_experiments/perf_drop_experiment_utils.py
code served (permissive licence) · get_code("cab47f390a020275")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned DataFrame. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import pandas as pd
import numpy as np

# Create a sample dataset with a split_variable column
np.random.seed(0)
n_rows = 2000
dataset_df = pd.DataFrame({
    'split_variable': np.random.choice(['domain_a', 'domain_b', 'domain_c'], n_rows),
    'feature_1': np.random.randn(n_rows),
    'feature_2': np.random.randn(n_rows),
    'feature_3': np.random.randn(n_rows),
    'target': np.random.randint(0, 2, n_rows)
})

ARGS = {
    'dataset_df': dataset_df,
    'split_variable': 'split_variable',
    'value': 'domain_a',
    'n_min_samples': 500,
    'n_max_samples': 10000,
    'seed': 0
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, pandas — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type DataFrame — a DataFrame output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a DataFrame output the recorded ndim 2 and dtype family float describe that coercion, not the output, and are not stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
sample_random_indices Ran
ran on a synthesized input
dataiku-research/performance_prediction_under_shift/drift_dac/perturbation_shared_utils.py
code served (permissive licence) · get_code("c945e0892ad1b1e4")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned list, ndim 1, dtype family int; no output shape is stated. The image it ran in was not recorded. If your copy returns the same type and ndim on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import numpy as np

# Set random seed for reproducibility
np.random.seed(42)

# Define arguments for sample_random_indices
ARGS = {
    "total_size": 100,
    "fraction": 0.3,
    "replace": False
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type list, ndim 1, dtype family int — a list output has no array shape of its own; none is stated
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
split_into_domains Ran
ran on a synthesized input
dataiku-research/performance_prediction_under_shift/drift_dac_experiments/perf_drop_experiment_utils.py
code served (permissive licence) · get_code("d18b5cfe530340f2")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned list, ndim 2, dtype family float; no output shape is stated. The image it ran in was not recorded. If your copy returns the same type and ndim on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import pandas as pd
import numpy as np

# Create a sample dataset with a split variable
np.random.seed(0)

# Create a dataframe with multiple domains
n_rows = 2000
data = {
    'feature_1': np.random.randn(n_rows),
    'feature_2': np.random.randn(n_rows),
    'feature_3': np.random.randn(n_rows),
    'domain': np.random.choice(['domain_A', 'domain_B', 'domain_C'], n_rows)
}

dataset_df = pd.DataFrame(data)

ARGS = {
    'dataset_df': dataset_df,
    'split_variable': 'domain',
    'n_min_samples': 500,
    'n_max_samples': 10000,
    'seed': 0
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, pandas — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type list, ndim 2, dtype family float — a list output has no array shape of its own; none is stated
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
subsample_feature_shift Ran
ran on a synthesized input
dataiku-research/performance_prediction_under_shift/drift_dac/covariate_shift.py
code served (permissive licence) · get_code("b32f1cda2af5ba27")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned tuple. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import numpy as np

# Set random seed for reproducibility
np.random.seed(0)

# Create small sample data
# x: feature matrix with 5 samples and 3 features
x = np.array([
    [1.0, 2.0, 3.0],
    [2.0, 3.0, 4.0],
    [3.0, 4.0, 5.0],
    [4.0, 5.0, 6.0],
    [5.0, 6.0, 7.0]
], dtype=np.float32)

# y: target vector with 5 samples
y = np.array([0, 1, 0, 1, 0], dtype=np.int32)

# Define arguments for subsample_feature_shift
ARGS = {
    'x': x,
    'y': y,
    'feat_delta': 0.5,
    'p': 0.5,
    'one_side': True
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type tuple — a tuple output has no array shape of its own; none is stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
subsample_one_feature_shift Ran
ran on a synthesized input
dataiku-research/performance_prediction_under_shift/drift_dac/covariate_shift.py
code served (permissive licence) · get_code("bbe9076675db4cbe")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned tuple. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import numpy as np

# Set random seed for reproducibility
np.random.seed(0)

# Create small sample data
x = np.array([
    [1.0, 2.0, 3.0, 4.0],
    [2.0, 3.0, 4.0, 5.0],
    [3.0, 4.0, 5.0, 6.0],
    [4.0, 5.0, 6.0, 7.0],
    [5.0, 6.0, 7.0, 8.0],
    [6.0, 7.0, 8.0, 9.0],
], dtype=np.float32)

y = np.array([0, 1, 0, 1, 0, 1], dtype=np.int32)

f = 0  # feature index
f_split = 3.5  # split threshold for feature 0

ARGS = {
    'x': x,
    'y': y,
    'f': f,
    'f_split': f_split,
    'p': 0.5,
    'one_side': True,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type tuple — a tuple output has no array shape of its own; none is stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
test_chi2 Ran
ran on a synthesized input
dataiku-research/performance_prediction_under_shift/drift_dac/drift_metrics.py
code served (permissive licence) · get_code("372e7aa065fc8cf5")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned tuple, ndim 1, dtype family float; no output shape is stated. The image it ran in was not recorded. If your copy returns the same type and ndim on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import numpy as np
import random

# Set random seed for reproducibility
random.seed(0)
np.random.seed(0)

# Create sample prediction arrays
# y_pred_src: source domain predictions (class indices)
# y_pred_tar: target domain predictions (class indices)
# nb_classes: number of classes

y_pred_src = np.array([0, 1, 2, 0, 1, 2, 0, 1, 2, 0], dtype=int)
y_pred_tar = np.array([0, 1, 2, 0, 1, 2, 0, 1], dtype=int)
nb_classes = 3

ARGS = {
    'y_pred_src': y_pred_src,
    'y_pred_tar': y_pred_tar,
    'nb_classes': nb_classes,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, scipy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type tuple, ndim 1, dtype family float — a tuple output has no array shape of its own; none is stated
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
univariate_drift Ran
ran on a synthesized input
dataiku-research/performance_prediction_under_shift/drift_dac/drift_metrics.py
code served (permissive licence) · get_code("8b37b6c06ce6a1bf")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned tuple, ndim 2, dtype family float; no output shape is stated. The image it ran in was not recorded. If your copy returns the same type and ndim on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import numpy as np
import scipy.stats as stats
import random

# Set seed for reproducibility
random.seed(0)
np.random.seed(0)

# Create small synthetic datasets for testing
# X_src: source domain data with 2 samples and 4 features
X_src = np.array([
    [1.5, 'cat_a', 2.3, 'group_1'],
    [2.1, 'cat_b', 1.8, 'group_2']
], dtype=object)

# X_tar: target domain data with 2 samples and 4 features
X_tar = np.array([
    [1.6, 'cat_a', 2.2, 'group_1'],
    [2.0, 'cat_b', 1.9, 'group_2']
], dtype=object)

# Convert numeric columns to float for proper processing
X_src_float = X_src.copy().astype(object)
X_src_float[:, 0] = X_src_float[:, 0].astype(float)
X_src_float[:, 2] = X_src_float[:, 2].astype(float)

X_tar_float = X_tar.copy().astype(object)
X_tar_float[:, 0] = X_tar_float[:, 0].astype(float)
X_tar_float[:, 2] = X_tar_float[:, 2].astype(float)

# is_categorical: boolean array indicating which features are categorical
# Feature 0: numeric, Feature 1: categorical, Feature 2: numeric, Feature 3: categorical
is_categorical = [False, True, False, True]

ARGS = {
    'X_src': X_src_float,
    'X_tar': X_tar_float,
    'is_categorical': is_categorical
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, scipy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type tuple, ndim 2, dtype family float — a tuple output has no array shape of its own; none is stated
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
is_integer_feature Not yet run
unverified (harvested, no recorded run)
dataiku-research/performance_prediction_under_shift/drift_dac/features_utils.py
code served (permissive licence) · get_code("d13b27c3bade1017")
How it ran

no invocation recorded.

subsample_joint_shift Not yet run
unverified (harvested, no recorded run)
dataiku-research/performance_prediction_under_shift/drift_dac/covariate_shift.py
code served (permissive licence) · get_code("41370631ad5ffc8a")
How it ran

no invocation recorded.

Also present in other papers' code

Of 16 samples lifted from this paper's linked repositories, 16 carry a function identity and 0 do not. Of the 16 distinct functions with an identity, 0 are also present in other papers whose code we hold (0 with at least one paper dated earlier, 0 the same month, 0 later) and 16 are only here among the 52,977 papers whose code Syntology holds (as of 2026-09-26).

Only here means: among the 52,977 papers whose code Syntology holds (51,887 with lifted samples, 1,090 through their assemblies only), as of 2026-09-26, no other paper's lifted samples or current assembly bricks contain an identical function body after normalisation.

Presence, not direction: an identical function body in two papers' repositories does not say who wrote it, which came from which, or that either paper's method rests on it. Matching is exact after normalisation, so a renamed or edited function is not matched. Dates are when papers appeared, not when code was written. Only .py files the harvester lifted are covered; notebooks are not read. A new-style arXiv id YYMM.NNNN dates a paper to the month 20YY-MM; otherwise Paper.year dates it to the year; otherwise it is undated. Dates are when the paper appeared, not when the code was written. Read 2026-11-22.

Repositories linked to this paper

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

Abstract

ML models deployed in production often have to face unknown domain changes, fundamentally different from their training settings. Performance prediction models carry out the crucial task of measuring the impact of these changes on model performance. We study the generalization capabilities of various performance prediction models to new domains by learning on generated synthetic perturbations. Empirical validation on a benchmark of ten tabular datasets shows that models based upon state-of-the-art shift detection metrics are not expressive enough to generalize to unseen domains, while Error Predictors bring a consistent improvement in performance prediction under shift. We additionally propose a natural and effortless uncertainty estimation of the predicted accuracy that ensures reliable use of performance predictors. Our implementation is available at https: //github.com/dataiku-research/performance_prediction_under_shift.

For agents

The same record, over MCP at https://syntology.ai/mcp:

get_harvested_code_for_paper("2206.10697")
get_code_for_paper("2206.10697")
have("2206.10697")

The run record, dated, one paper per request, free:

curl https://syntology.ai/api/ran/2206.10697.json

A badge for a README (the split and the date, never a ratio):

[![Syntology run record](https://syntology.ai/api/ran/2206.10697.svg)](https://syntology.ai/paper/2206.10697)

Connect an agent — have() is free.