SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2108.07927 · 2021

Fed-TGAN: Federated Learning Framework for Synthesizing Tabular Data

arXiv · PDF · Open in the Atlas

Code that ran

8 of 8 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 8 have no contract verdict; 0 failures of our instrument; 0 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/2108.07927.json · badge: /api/ran/2108.07927.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
smduan/Fed-TDA community (classifier label)
classifier basis not recorded on the link
8 ran · 0 no recorded run
FunctionStatusWhere it lives
fed_max Ran
ran on a synthesized input
smduan/Fed-TDA/fedtda/fed_compute.py
code served (permissive licence) · get_code("2180516c896796db")
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 pandas as pd

# Test case 1: list of pandas Series
series1 = pd.Series([1, 5, 3, 2])
series2 = pd.Series([10, 2, 8, 4])
series3 = pd.Series([3, 7, 1, 6])

max_list_series = [series1, series2, series3]

ARGS = {
    "max_list": max_list_series
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
pandas — 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
fed_mean Ran
ran on a synthesized input
smduan/Fed-TDA/fedtda/fed_compute.py
code served (permissive licence) · get_code("4a210af80a442128")
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 float64, ndim 0, 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

torch_manual_seed = 0
np.random.seed(0)

# The function fed_mean takes x_list, which should be a list of array-like objects
# that have .sum() method and support len()
# Looking at the function:
# - It iterates over x_list
# - For each x, it calls x.sum() and len(x)
# - It computes the mean across all elements

# Create small numpy arrays with distinct sizes to make dimensions unambiguous
x1 = np.array([[1.0, 2.0, 3.0, 4.0],
               [5.0, 6.0, 7.0, 8.0]])  # shape (2, 4)

x2 = np.array([[9.0, 10.0, 11.0, 12.0],
               [13.0, 14.0, 15.0, 16.0],
               [17.0, 18.0, 19.0, 20.0]])  # shape (3, 4)

x_list = [x1, x2]

ARGS = {
    "x_list": x_list
}
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 float64, ndim 0, dtype family float — no output shape is stated: observe_signature.py recorded float ndim 0, which disagrees with the fixture run
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
fed_min Ran
ran on a synthesized input
smduan/Fed-TDA/fedtda/fed_compute.py
code served (permissive licence) · get_code("a76e237b4869e118")
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 float64, ndim 0, 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 torch

# Case 1: min_list contains pd.Series objects
series1 = pd.Series([1, 2, 3, 4, 5])
series2 = pd.Series([0.5, 1.5, 2.5])
series3 = pd.Series([10, 20, 30, 40])

min_list_series = [series1, series2, series3]

ARGS = {
    "min_list": min_list_series
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
pandas, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type float64, ndim 0, dtype family float — no output shape is stated: observe_signature.py recorded float ndim 0, which disagrees with the fixture run
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
fed_value_counts Ran
ran on a synthesized input
smduan/Fed-TDA/fedtda/CategoricalTransformer.py
code served (permissive licence) · get_code("5b0a132336560741")
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 Series. 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 sample discrete data as pandas Series objects
# The function calls .value_counts(dropna=False) on each item in data_list
# and accesses .index on the resulting Series

series1 = pd.Series(['A', 'B', 'A', 'C', 'B', 'A', None])
series2 = pd.Series(['B', 'C', 'C', 'D', 'A', None])
series3 = pd.Series(['A', 'D', 'E', 'B'])

data_list = [series1, series2, series3]

ARGS = {
    'data_list': data_list
}
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 Series — a Series output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a Series output the recorded ndim 1 and dtype family int 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
get_clients_distribution Ran
ran on a synthesized input
smduan/Fed-TDA/clinical_TDA_syn.py
code served (permissive licence) · get_code("7fe005c438a5def8")
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 Counter. 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 torch
import numpy as np
from collections import Counter

torch.manual_seed(0)

# Construct datasets as a list of dictionaries
# Each dataset dict has a 'label' key mapping to a list of label indices
datasets = [
    {'label': [0, 1, 0, 2, 1]},
    {'label': [1, 2, 0, 0, 1]},
    {'label': [2, 2, 1, 0, 0]}
]

label_name = 'label'
label_num = 3

ARGS = {
    'datasets': datasets,
    'label_name': label_name,
    'label_num': label_num
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Counter — a Counter output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a Counter output the recorded ndim 0 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
get_dataset Ran
ran on a synthesized input
smduan/Fed-TDA/fedavg/datasets.py
code served (permissive licence) · get_code("b59e926532eac625")
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 MyTabularDataset. 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 torch
import pandas as pd
from PIL import Image
import io

torch.manual_seed(0)

# Create a small tabular dataset
tabular_data = pd.DataFrame({
    'feature1': [1.0, 2.0],
    'feature2': [3.0, 4.0],
    'feature3': [5.0, 6.0],
    'label': [0, 1]
})

# Create a small image dataset with fake in-memory image data
def create_fake_image():
    img = Image.new('RGB', (8, 8), color=(73, 109, 137))
    return img

# Create temporary image files in memory using PIL
image_data = []
for i in range(2):
    img = create_fake_image()
    img_bytes = io.BytesIO()
    img.save(img_bytes, format='PNG')
    img_bytes.seek(0)
    image_data.append(img_bytes)

# Since we can't use real file paths, we'll create a mock dataset
# that the function can work with for tabular case
image_dataset = pd.DataFrame({
    'image_path': ['dummy_path_0.png', 'dummy_path_1.png'],
    'label': [0, 1]
})

# Configuration for tabular dataset
conf_tabular = {
    'data_type': 'tabular',
    'label_column': 'label'
}

# Configuration for image dataset (won't actually load files in this test)
conf_image = {
    'data_type': 'image',
    'data_column': 'image_path',
    'label_column': 'label'
}

ARGS = {
    'conf': conf_tabular,
    'data': tabular_data
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
PIL, pandas, torch, torchvision — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type MyTabularDataset — a MyTabularDataset 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
label_skew Ran
ran on a synthesized input
smduan/Fed-TDA/utils.py
code served (permissive licence) · get_code("f19aec409d579b3d")
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 small dataset with labels
np.random.seed(0)
n_samples = 100
n_features = 4
n_classes = 3

# Create synthetic data
X = np.random.randn(n_samples, n_features)
y = np.random.randint(0, n_classes, n_samples)

# Create a DataFrame
data = pd.DataFrame(X, columns=[f'feature_{i}' for i in range(n_features)])
data['label'] = y

ARGS = {
    'data': data,
    'label': 'label',
    'K': 3,
    'n_parties': 4,
    'beta': 0.5,
    'min_require_size': 10
}
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
read_train_datasets Ran
ran on a synthesized input
smduan/Fed-TDA/clinical_TDA_syn.py
code served (permissive licence) · get_code("6a323d8e0bfa32aa")
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 3, 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 tempfile
import os

# Create temporary directory with test CSV files
temp_dir = tempfile.mkdtemp()

# Create sample CSV files matching the expected naming pattern
label_name = "label"
client_num = 3

for i in range(client_num):
    df = pd.DataFrame({
        'feature1': [1.0, 2.0],
        'feature2': [3.0, 4.0],
        label_name: [0, 1]
    })
    filepath = os.path.join(temp_dir, f'{label_name}_{i}.csv')
    df.to_csv(filepath, index=False)

# Define arguments dict
ARGS = {
    'path': temp_dir + '/',
    'client_num': client_num,
    'label_name': label_name
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
pandas — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type list, ndim 3, 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

Also present in other papers' code

Of 8 samples lifted from this paper's linked repositories, 8 carry a function identity and 0 do not. Of the 8 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 8 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

Generative Adversarial Networks (GANs) are typically trained to synthesize data, from images and more recently tabular data, under the assumption of directly accessible training data. Recently, federated learning (FL) is an emerging paradigm that features decentralized learning on client's local data with a privacy-preserving capability. And, while learning GANs to synthesize images on FL systems has just been demonstrated, it is unknown if GANs for tabular data can be learned from decentralized data sources. Moreover, it remains unclear which distributed architecture suits them best. Different from image GANs, state-of-the-art tabular GANs require prior knowledge on the data distribution of each (discrete and continuous) column to agree on a common encoding -- risking privacy guarantees. In this paper, we propose Fed-TGAN, the first Federated learning framework for Tabular GANs. To effectively learn a complex tabular GAN on non-identical participants, Fed-TGAN designs two novel features: (i) a privacy-preserving multi-source feature encoding for model initialization; and (ii) table similarity aware weighting strategies to aggregate local models for countering data skew. We extensively evaluate the proposed Fed-TGAN against variants of decentralized learning architectures on four widely used datasets. Results show that Fed-TGAN accelerates training time per epoch up to 200% compared to the alternative architectures, for both IID and Non-IID data. Overall, Fed-TGAN not only stabilizes the training loss, but also achieves better similarity between generated and original data. Our code is released at https://github.com/zhao-zilong/Fed-TGAN.

For agents

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

get_harvested_code_for_paper("2108.07927")
get_code_for_paper("2108.07927")
have("2108.07927")

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

curl https://syntology.ai/api/ran/2108.07927.json

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

[![Syntology run record](https://syntology.ai/api/ran/2108.07927.svg)](https://syntology.ai/paper/2108.07927)

Connect an agent — have() is free.