kmodel 0.0.1__tar.gz

kmodel-0.0.1/LICENSE

@@ -0,0 +1,201 @@
+                                 Apache License
+                           Version 2.0, January 2004
+                        http://www.apache.org/licenses/
+
+   TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION
+
+   1. Definitions.
+
+      "License" shall mean the terms and conditions for use, reproduction,
+      and distribution as defined by Sections 1 through 9 of this document.
+
+      "Licensor" shall mean the copyright owner or entity authorized by
+      the copyright owner that is granting the License.
+
+      "Legal Entity" shall mean the union of the acting entity and all
+      other entities that control, are controlled by, or are under common
+      control with that entity. For the purposes of this definition,
+      "control" means (i) the power, direct or indirect, to cause the
+      direction or management of such entity, whether by contract or
+      otherwise, or (ii) ownership of fifty percent (50%) or more of the
+      outstanding shares, or (iii) beneficial ownership of such entity.
+
+      "You" (or "Your") shall mean an individual or Legal Entity
+      exercising permissions granted by this License.
+
+      "Source" form shall mean the preferred form for making modifications,
+      including but not limited to software source code, documentation
+      source, and configuration files.
+
+      "Object" form shall mean any form resulting from mechanical
+      transformation or translation of a Source form, including but
+      not limited to compiled object code, generated documentation,
+      and conversions to other media types.
+
+      "Work" shall mean the work of authorship, whether in Source or
+      Object form, made available under the License, as indicated by a
+      copyright notice that is included in or attached to the work
+      (an example is provided in the Appendix below).
+
+      "Derivative Works" shall mean any work, whether in Source or Object
+      form, that is based on (or derived from) the Work and for which the
+      editorial revisions, annotations, elaborations, or other modifications
+      represent, as a whole, an original work of authorship. For the purposes
+      of this License, Derivative Works shall not include works that remain
+      separable from, or merely link (or bind by name) to the interfaces of,
+      the Work and Derivative Works thereof.
+
+      "Contribution" shall mean any work of authorship, including
+      the original version of the Work and any modifications or additions
+      to that Work or Derivative Works thereof, that is intentionally
+      submitted to Licensor for inclusion in the Work by the copyright owner
+      or by an individual or Legal Entity authorized to submit on behalf of
+      the copyright owner. For the purposes of this definition, "submitted"
+      means any form of electronic, verbal, or written communication sent
+      to the Licensor or its representatives, including but not limited to
+      communication on electronic mailing lists, source code control systems,
+      and issue tracking systems that are managed by, or on behalf of, the
+      Licensor for the purpose of discussing and improving the Work, but
+      excluding communication that is conspicuously marked or otherwise
+      designated in writing by the copyright owner as "Not a Contribution."
+
+      "Contributor" shall mean Licensor and any individual or Legal Entity
+      on behalf of whom a Contribution has been received by Licensor and
+      subsequently incorporated within the Work.
+
+   2. Grant of Copyright License. Subject to the terms and conditions of
+      this License, each Contributor hereby grants to You a perpetual,
+      worldwide, non-exclusive, no-charge, royalty-free, irrevocable
+      copyright license to reproduce, prepare Derivative Works of,
+      publicly display, publicly perform, sublicense, and distribute the
+      Work and such Derivative Works in Source or Object form.
+
+   3. Grant of Patent License. Subject to the terms and conditions of
+      this License, each Contributor hereby grants to You a perpetual,
+      worldwide, non-exclusive, no-charge, royalty-free, irrevocable
+      (except as stated in this section) patent license to make, have made,
+      use, offer to sell, sell, import, and otherwise transfer the Work,
+      where such license applies only to those patent claims licensable
+      by such Contributor that are necessarily infringed by their
+      Contribution(s) alone or by combination of their Contribution(s)
+      with the Work to which such Contribution(s) was submitted. If You
+      institute patent litigation against any entity (including a
+      cross-claim or counterclaim in a lawsuit) alleging that the Work
+      or a Contribution incorporated within the Work constitutes direct
+      or contributory patent infringement, then any patent licenses
+      granted to You under this License for that Work shall terminate
+      as of the date such litigation is filed.
+
+   4. Redistribution. You may reproduce and distribute copies of the
+      Work or Derivative Works thereof in any medium, with or without
+      modifications, and in Source or Object form, provided that You
+      meet the following conditions:
+
+      (a) You must give any other recipients of the Work or
+          Derivative Works a copy of this License; and
+
+      (b) You must cause any modified files to carry prominent notices
+          stating that You changed the files; and
+
+      (c) You must retain, in the Source form of any Derivative Works
+          that You distribute, all copyright, patent, trademark, and
+          attribution notices from the Source form of the Work,
+          excluding those notices that do not pertain to any part of
+          the Derivative Works; and
+
+      (d) If the Work includes a "NOTICE" text file as part of its
+          distribution, then any Derivative Works that You distribute must
+          include a readable copy of the attribution notices contained
+          within such NOTICE file, excluding those notices that do not
+          pertain to any part of the Derivative Works, in at least one
+          of the following places: within a NOTICE text file distributed
+          as part of the Derivative Works; within the Source form or
+          documentation, if provided along with the Derivative Works; or,
+          within a display generated by the Derivative Works, if and
+          wherever such third-party notices normally appear. The contents
+          of the NOTICE file are for informational purposes only and
+          do not modify the License. You may add Your own attribution
+          notices within Derivative Works that You distribute, alongside
+          or as an addendum to the NOTICE text from the Work, provided
+          that such additional attribution notices cannot be construed
+          as modifying the License.
+
+      You may add Your own copyright statement to Your modifications and
+      may provide additional or different license terms and conditions
+      for use, reproduction, or distribution of Your modifications, or
+      for any such Derivative Works as a whole, provided Your use,
+      reproduction, and distribution of the Work otherwise complies with
+      the conditions stated in this License.
+
+   5. Submission of Contributions. Unless You explicitly state otherwise,
+      any Contribution intentionally submitted for inclusion in the Work
+      by You to the Licensor shall be under the terms and conditions of
+      this License, without any additional terms or conditions.
+      Notwithstanding the above, nothing herein shall supersede or modify
+      the terms of any separate license agreement you may have executed
+      with Licensor regarding such Contributions.
+
+   6. Trademarks. This License does not grant permission to use the trade
+      names, trademarks, service marks, or product names of the Licensor,
+      except as required for reasonable and customary use in describing the
+      origin of the Work and reproducing the content of the NOTICE file.
+
+   7. Disclaimer of Warranty. Unless required by applicable law or
+      agreed to in writing, Licensor provides the Work (and each
+      Contributor provides its Contributions) on an "AS IS" BASIS,
+      WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
+      implied, including, without limitation, any warranties or conditions
+      of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A
+      PARTICULAR PURPOSE. You are solely responsible for determining the
+      appropriateness of using or redistributing the Work and assume any
+      risks associated with Your exercise of permissions under this License.
+
+   8. Limitation of Liability. In no event and under no legal theory,
+      whether in tort (including negligence), contract, or otherwise,
+      unless required by applicable law (such as deliberate and grossly
+      negligent acts) or agreed to in writing, shall any Contributor be
+      liable to You for damages, including any direct, indirect, special,
+      incidental, or consequential damages of any character arising as a
+      result of this License or out of the use or inability to use the
+      Work (including but not limited to damages for loss of goodwill,
+      work stoppage, computer failure or malfunction, or any and all
+      other commercial damages or losses), even if such Contributor
+      has been advised of the possibility of such damages.
+
+   9. Accepting Warranty or Additional Liability. While redistributing
+      the Work or Derivative Works thereof, You may choose to offer,
+      and charge a fee for, acceptance of support, warranty, indemnity,
+      or other liability obligations and/or rights consistent with this
+      License. However, in accepting such obligations, You may act only
+      on Your own behalf and on Your sole responsibility, not on behalf
+      of any other Contributor, and only if You agree to indemnify,
+      defend, and hold each Contributor harmless for any liability
+      incurred by, or claims asserted against, such Contributor by reason
+      of your accepting any such warranty or additional liability.
+
+   END OF TERMS AND CONDITIONS
+
+   APPENDIX: How to apply the Apache License to your work.
+
+      To apply the Apache License to your work, attach the following
+      boilerplate notice, with the fields enclosed by brackets "[]"
+      replaced with your own identifying information. (Don't include
+      the brackets!)  The text should be enclosed in the appropriate
+      comment syntax for the file format. We also recommend that a
+      file or class name and description of purpose be included on the
+      same "printed page" as the copyright notice for easier
+      identification within third-party archives.
+
+   Copyright [yyyy] [name of copyright owner]
+
+   Licensed under the Apache License, Version 2.0 (the "License");
+   you may not use this file except in compliance with the License.
+   You may obtain a copy of the License at
+
+       http://www.apache.org/licenses/LICENSE-2.0
+
+   Unless required by applicable law or agreed to in writing, software
+   distributed under the License is distributed on an "AS IS" BASIS,
+   WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+   See the License for the specific language governing permissions and
+   limitations under the License.

kmodel-0.0.1/MANIFEST.in

@@ -0,0 +1,5 @@
+include pyproject.toml
+include LICENSE
+include CONTRIBUTING.md
+include README.md
+recursive-exclude * __pycache__

kmodel-0.0.1/PKG-INFO

@@ -0,0 +1,89 @@
+Metadata-Version: 2.4
+Name: kmodel
+Version: 0.0.1
+Summary: A fresh nbdev project for kmodel.
+Author-email: keke <lcai888666@gmail.com>
+License: Apache-2.0
+Project-URL: Repository, https://github.com/sky1ove/kmodel
+Project-URL: Documentation, https://sky1ove.github.io/kmodel/
+Keywords: nbdev
+Classifier: Programming Language :: Python :: 3
+Classifier: Programming Language :: Python :: 3 :: Only
+Requires-Python: >=3.10
+Description-Content-Type: text/markdown
+License-File: LICENSE
+Requires-Dist: fastai>=2.8.7
+Requires-Dist: joblib>=1.5.3
+Requires-Dist: numpy>=2.2.6
+Requires-Dist: pandas>=2.3.3
+Requires-Dist: scikit-learn>=1.7.2
+Requires-Dist: torch>=2.11.0
+Dynamic: license-file
+
+# kmodel
+
+
+<!-- WARNING: THIS FILE WAS AUTOGENERATED! DO NOT EDIT! -->
+
+This file will become your README and also the index of your
+documentation.
+
+## Developer Guide
+
+If you are new to using `nbdev` here are some useful pointers to get you
+started.
+
+### Install kmodel in Development mode
+
+``` sh
+# make sure kmodel package is installed in development mode
+$ pip install -e .
+
+# make changes under nbs/ directory
+# ...
+
+# compile to have changes apply to kmodel
+$ nbdev_prepare
+```
+
+## Usage
+
+### Installation
+
+Install latest from the GitHub
+[repository](https://github.com/sky1ove/kmodel):
+
+``` sh
+$ pip install git+https://github.com/sky1ove/kmodel.git
+```
+
+or from [conda](https://anaconda.org/sky1ove/kmodel)
+
+``` sh
+$ conda install -c sky1ove kmodel
+```
+
+or from [pypi](https://pypi.org/project/kmodel/)
+
+``` sh
+$ pip install kmodel
+```
+
+### Documentation
+
+Documentation can be found hosted on this GitHub
+[repository](https://github.com/sky1ove/kmodel)’s
+[pages](https://sky1ove.github.io/kmodel/). Additionally you can find
+package manager specific guidelines on
+[conda](https://anaconda.org/sky1ove/kmodel) and
+[pypi](https://pypi.org/project/kmodel/) respectively.
+
+## How to use
+
+Fill me in please! Don’t forget code examples:
+
+``` python
+1+1
+```
+
+    2

kmodel-0.0.1/README.md

@@ -0,0 +1,67 @@
+# kmodel
+
+
+<!-- WARNING: THIS FILE WAS AUTOGENERATED! DO NOT EDIT! -->
+
+This file will become your README and also the index of your
+documentation.
+
+## Developer Guide
+
+If you are new to using `nbdev` here are some useful pointers to get you
+started.
+
+### Install kmodel in Development mode
+
+``` sh
+# make sure kmodel package is installed in development mode
+$ pip install -e .
+
+# make changes under nbs/ directory
+# ...
+
+# compile to have changes apply to kmodel
+$ nbdev_prepare
+```
+
+## Usage
+
+### Installation
+
+Install latest from the GitHub
+[repository](https://github.com/sky1ove/kmodel):
+
+``` sh
+$ pip install git+https://github.com/sky1ove/kmodel.git
+```
+
+or from [conda](https://anaconda.org/sky1ove/kmodel)
+
+``` sh
+$ conda install -c sky1ove kmodel
+```
+
+or from [pypi](https://pypi.org/project/kmodel/)
+
+``` sh
+$ pip install kmodel
+```
+
+### Documentation
+
+Documentation can be found hosted on this GitHub
+[repository](https://github.com/sky1ove/kmodel)’s
+[pages](https://sky1ove.github.io/kmodel/). Additionally you can find
+package manager specific guidelines on
+[conda](https://anaconda.org/sky1ove/kmodel) and
+[pypi](https://pypi.org/project/kmodel/) respectively.
+
+## How to use
+
+Fill me in please! Don’t forget code examples:
+
+``` python
+1+1
+```
+
+    2

kmodel-0.0.1/kmodel/__init__.py

@@ -0,0 +1 @@
+__version__ = "0.0.1"

kmodel-0.0.1/kmodel/_modidx.py

@@ -0,0 +1,41 @@
+# Autogenerated by nbdev
+
+d = { 'settings': { 'branch': 'main',
+                'doc_baseurl': '/kmodel',
+                'doc_host': 'https://sky1ove.github.io',
+                'git_url': 'https://github.com/sky1ove/kmodel',
+                'lib_path': 'kmodel'},
+  'syms': { 'kmodel.core': {'kmodel.core.foo': ('core.html#foo', 'kmodel/core.py')},
+            'kmodel.dnn': { 'kmodel.dnn.CE': ('dnn.html#ce', 'kmodel/dnn.py'),
+                            'kmodel.dnn.CNN1D': ('dnn.html#cnn1d', 'kmodel/dnn.py'),
+                            'kmodel.dnn.CNN1D.__init__': ('dnn.html#cnn1d.__init__', 'kmodel/dnn.py'),
+                            'kmodel.dnn.CNN1D.forward': ('dnn.html#cnn1d.forward', 'kmodel/dnn.py'),
+                            'kmodel.dnn.GeneralDataset': ('dnn.html#generaldataset', 'kmodel/dnn.py'),
+                            'kmodel.dnn.GeneralDataset.__getitem__': ('dnn.html#generaldataset.__getitem__', 'kmodel/dnn.py'),
+                            'kmodel.dnn.GeneralDataset.__init__': ('dnn.html#generaldataset.__init__', 'kmodel/dnn.py'),
+                            'kmodel.dnn.GeneralDataset.__len__': ('dnn.html#generaldataset.__len__', 'kmodel/dnn.py'),
+                            'kmodel.dnn.JSD': ('dnn.html#jsd', 'kmodel/dnn.py'),
+                            'kmodel.dnn.KLD': ('dnn.html#kld', 'kmodel/dnn.py'),
+                            'kmodel.dnn.MLP': ('dnn.html#mlp', 'kmodel/dnn.py'),
+                            'kmodel.dnn.PSSM_model': ('dnn.html#pssm_model', 'kmodel/dnn.py'),
+                            'kmodel.dnn.PSSM_model.__init__': ('dnn.html#pssm_model.__init__', 'kmodel/dnn.py'),
+                            'kmodel.dnn.PSSM_model.forward': ('dnn.html#pssm_model.forward', 'kmodel/dnn.py'),
+                            'kmodel.dnn.conv_wn': ('dnn.html#conv_wn', 'kmodel/dnn.py'),
+                            'kmodel.dnn.init_weights': ('dnn.html#init_weights', 'kmodel/dnn.py'),
+                            'kmodel.dnn.lin_wn': ('dnn.html#lin_wn', 'kmodel/dnn.py'),
+                            'kmodel.dnn.predict_dl': ('dnn.html#predict_dl', 'kmodel/dnn.py'),
+                            'kmodel.dnn.seed_everything': ('dnn.html#seed_everything', 'kmodel/dnn.py'),
+                            'kmodel.dnn.train_dl': ('dnn.html#train_dl', 'kmodel/dnn.py'),
+                            'kmodel.dnn.train_dl_cv': ('dnn.html#train_dl_cv', 'kmodel/dnn.py')},
+            'kmodel.train': { 'kmodel.train._normalize_vector': ('ml.html#_normalize_vector', 'kmodel/train.py'),
+                              'kmodel.train.calculate_ce': ('ml.html#calculate_ce', 'kmodel/train.py'),
+                              'kmodel.train.get_score': ('ml.html#get_score', 'kmodel/train.py'),
+                              'kmodel.train.get_splits': ('ml.html#get_splits', 'kmodel/train.py'),
+                              'kmodel.train.js_divergence_flat': ('ml.html#js_divergence_flat', 'kmodel/train.py'),
+                              'kmodel.train.kl_divergence_flat': ('ml.html#kl_divergence_flat', 'kmodel/train.py'),
+                              'kmodel.train.post_process': ('ml.html#post_process', 'kmodel/train.py'),
+                              'kmodel.train.post_process_oof': ('ml.html#post_process_oof', 'kmodel/train.py'),
+                              'kmodel.train.predict_ml': ('ml.html#predict_ml', 'kmodel/train.py'),
+                              'kmodel.train.split_data': ('ml.html#split_data', 'kmodel/train.py'),
+                              'kmodel.train.train_ml': ('ml.html#train_ml', 'kmodel/train.py'),
+                              'kmodel.train.train_ml_cv': ('ml.html#train_ml_cv', 'kmodel/train.py')}}}

kmodel-0.0.1/kmodel/core.py

@@ -0,0 +1,9 @@
+"""Fill in a module description here"""
+
+# AUTOGENERATED! DO NOT EDIT! File to edit: ../nbs/00_core.ipynb.
+
+# %% auto #0
+__all__ = ['foo']
+
+# %% ../nbs/00_core.ipynb #97ca5313
+def foo(): pass

kmodel-0.0.1/kmodel/dnn.py

@@ -0,0 +1,291 @@
+"""Deep learning helpers built on the existing fastai training backbone, with runnable examples from a seaborn dataset."""
+
+# AUTOGENERATED! DO NOT EDIT! File to edit: ../nbs/02_DNN.ipynb.
+
+# %% auto #0
+__all__ = ['def_device', 'seed_everything', 'GeneralDataset', 'MLP', 'lin_wn', 'conv_wn', 'CNN1D', 'init_weights', 'PSSM_model',
+           'CE', 'KLD', 'JSD', 'train_dl', 'predict_dl', 'train_dl_cv']
+
+# %% ../nbs/02_DNN.ipynb #19eb18c4
+import os
+import random
+
+import fastcore.all as fc
+import matplotlib.pyplot as plt
+import numpy as np
+import pandas as pd
+import torch
+import torch.nn as nn
+import torch.nn.functional as F
+from fastai.vision.all import *
+from torch.utils.data import DataLoader, Dataset
+
+# %% ../nbs/02_DNN.ipynb #43dceac1
+def seed_everything(seed: int = 123) -> None:
+    "Seed Python, NumPy, and PyTorch for reproducible experiments." 
+    random.seed(seed)
+    os.environ['PYTHONHASHSEED'] = str(seed)
+    np.random.seed(seed)
+    torch.manual_seed(seed)
+    if torch.cuda.is_available():
+        torch.cuda.manual_seed_all(seed)
+    torch.backends.cudnn.deterministic = True
+    torch.backends.cudnn.benchmark = False
+
+# %% ../nbs/02_DNN.ipynb #cc4d6625
+def_device = 'mps' if torch.backends.mps.is_available() else 'cuda' if torch.cuda.is_available() else 'cpu'
+
+# %% ../nbs/02_DNN.ipynb #e5a649d5
+class GeneralDataset(Dataset):
+    def __init__(
+        self,
+        df: pd.DataFrame,
+        feat_col,
+        target_col=None,
+        A: int = 23,
+        dtype=np.float32,
+    ) -> None:
+        "Return features only in test mode, or (features, targets) in training mode." 
+        self.test = target_col is None
+        self.aa = A
+        self.X = df[list(feat_col)].to_numpy(dtype=dtype, copy=True)
+        self.y = None
+        if not self.test:
+            y_flat = df[list(target_col)].to_numpy(dtype=dtype, copy=True)
+            total = y_flat.shape[1]
+            if total % A != 0:
+                raise ValueError(f"Target columns ({total}) not divisible by A={A}; cannot infer positions.")
+            self.position = total // self.aa
+            self.y = y_flat.reshape(-1, A, self.position)
+        self.len = len(df)
+
+    def __len__(self):
+        return self.len
+
+    def __getitem__(self, index):
+        X = torch.from_numpy(self.X[index])
+        if self.test:
+            return X
+        y = torch.from_numpy(self.y[index])
+        return X, y
+
+# %% ../nbs/02_DNN.ipynb #d358392f
+def MLP(
+    num_features: int,
+    num_targets: int,
+    hidden_units: list[int] = [512, 218],
+    dp: float = 0.2,
+):
+    "Feed-forward model for tabular inputs." 
+    layers = [
+        nn.Linear(num_features, hidden_units[0]),
+        nn.BatchNorm1d(hidden_units[0]),
+        nn.Dropout(dp),
+        nn.PReLU(),
+    ]
+    for i in range(len(hidden_units) - 1):
+        layers.extend([
+            nn.Linear(hidden_units[i], hidden_units[i + 1]),
+            nn.BatchNorm1d(hidden_units[i + 1]),
+            nn.Dropout(dp),
+            nn.PReLU(),
+        ])
+    layers.append(nn.Linear(hidden_units[-1], num_targets))
+    return nn.Sequential(*layers)
+
+# %% ../nbs/02_DNN.ipynb #59c86969
+def lin_wn(ni, nf, dp: float = 0.1, act=nn.SiLU):
+    "Weight-normalized linear block." 
+    layers = [
+        nn.BatchNorm1d(ni),
+        nn.Dropout(dp),
+        nn.utils.parametrizations.weight_norm(nn.Linear(ni, nf)),
+    ]
+    if act:
+        layers.append(act())
+    return nn.Sequential(*layers)
+
+# %% ../nbs/02_DNN.ipynb #25d0985c
+def conv_wn(ni, nf, ks: int = 3, stride: int = 1, padding: int = 1, dp: float = 0.1, act=nn.ReLU):
+    "Weight-normalized convolution block." 
+    layers = [
+        nn.BatchNorm1d(ni),
+        nn.Dropout(dp),
+        nn.utils.parametrizations.weight_norm(nn.Conv1d(ni, nf, ks, stride, padding)),
+    ]
+    if act:
+        layers.append(act())
+    return nn.Sequential(*layers)
+
+# %% ../nbs/02_DNN.ipynb #b6691f78
+class CNN1D(nn.Module):
+    def __init__(self, ni, nf, amp_scale: int = 16):
+        super().__init__()
+        cha_1, cha_2, cha_3 = 256, 512, 512
+        hidden_size = cha_1 * amp_scale
+        cha_po_1 = hidden_size // (cha_1 * 2)
+        cha_po_2 = (hidden_size // (cha_1 * 4)) * cha_3
+
+        self.lin = lin_wn(ni, hidden_size)
+        self.view = View(-1, cha_1, amp_scale)
+        self.conv1 = nn.Sequential(
+            conv_wn(cha_1, cha_2, ks=5, stride=1, padding=2, dp=0.1),
+            nn.AdaptiveAvgPool1d(output_size=cha_po_1),
+            conv_wn(cha_2, cha_2, ks=3, stride=1, padding=1, dp=0.1),
+        )
+        self.conv2 = nn.Sequential(
+            conv_wn(cha_2, cha_2, ks=3, stride=1, padding=1, dp=0.3),
+            conv_wn(cha_2, cha_3, ks=5, stride=1, padding=2, dp=0.2),
+        )
+        self.head = nn.Sequential(
+            nn.MaxPool1d(kernel_size=4, stride=2, padding=1),
+            nn.Flatten(),
+            lin_wn(cha_po_2, nf, act=None),
+        )
+
+    def forward(self, x):
+        x = self.lin(x)
+        x = self.view(x)
+        x = self.conv1(x)
+        x_s = x
+        x = self.conv2(x)
+        x = x * x_s
+        return self.head(x)
+
+# %% ../nbs/02_DNN.ipynb #32493367
+def init_weights(m, leaky: float = 0.0):
+    "Initialize convolution layers with Kaiming normal weights." 
+    if isinstance(m, (nn.Conv1d, nn.Conv2d, nn.Conv3d)):
+        nn.init.kaiming_normal_(m.weight, a=leaky)
+
+# %% ../nbs/02_DNN.ipynb #bad5feef
+class PSSM_model(nn.Module):
+    def __init__(self, n_features, n_targets, A: int = 23, model: str = 'MLP'):
+        super().__init__()
+        self.n_features = n_features
+        self.n_targets = n_targets
+        self.n_aa = A
+        if self.n_targets % self.n_aa != 0:
+            raise ValueError(f"n_targets ({n_targets}) must be divisible by n_aa ({self.n_aa}).")
+        self.n_positions = self.n_targets // self.n_aa
+
+        if model == 'MLP':
+            self.model = MLP(self.n_features, self.n_targets)
+        elif model == 'CNN':
+            self.model = CNN1D(self.n_features, self.n_targets).apply(init_weights)
+        else:
+            raise ValueError('model must be MLP or CNN.')
+
+    def forward(self, x):
+        logits = self.model(x).reshape(-1, self.n_aa, self.n_positions)
+        return logits
+
+# %% ../nbs/02_DNN.ipynb #f87f424e
+def CE(logits: torch.Tensor, target_probs: torch.Tensor):
+    "Cross-entropy with soft labels." 
+    logp = F.log_softmax(logits, dim=1)
+    ce = -(target_probs * logp).sum(dim=1)
+    return ce.mean()
+
+# %% ../nbs/02_DNN.ipynb #ccb32c56
+def KLD(logits: torch.Tensor, target_probs: torch.Tensor):
+    "Average KL divergence across positions between target_probs and softmax(logits)." 
+    logq = F.log_softmax(logits, dim=1)
+    logp = torch.log(target_probs + 1e-8)
+    kl = (target_probs * (logp - logq)).sum(dim=1)
+    return kl.mean()
+
+# %% ../nbs/02_DNN.ipynb #fe899c20
+def JSD(logits: torch.Tensor, target_probs: torch.Tensor):
+    "Average Jensen-Shannon divergence across positions between target_probs and softmax(logits)." 
+    q = F.softmax(logits, dim=1)
+    p = target_probs
+    m = 0.5 * (p + q)
+    logp = torch.log(p + 1e-8)
+    logq = torch.log(q + 1e-8)
+    logm = torch.log(m + 1e-8)
+    kld_pm = (p * (logp - logm)).sum(dim=1)
+    kld_qm = (q * (logq - logm)).sum(dim=1)
+    return (0.5 * (kld_pm + kld_qm)).mean()
+
+# %% ../nbs/02_DNN.ipynb #d0f14083
+def train_dl(
+    df: pd.DataFrame,
+    feat_col,
+    target_col,
+    split,
+    model_func,
+    A: int = 23,
+    n_epoch: int = 4,
+    bs: int = 32,
+    lr: float = 1e-2,
+    loss=CE,
+    save=None,
+    sampler=None,
+    lr_find: bool = False,
+):
+    "Train a deep learning model with the fastai learner stack." 
+    train = df.iloc[split[0]]
+    valid = df.iloc[split[1]]
+
+    train_ds = GeneralDataset(train, feat_col, target_col, A=A)
+    valid_ds = GeneralDataset(valid, feat_col, target_col, A=A)
+    dls = DataLoaders.from_dsets(train_ds, valid_ds, bs=bs, num_workers=min(fc.defaults.cpus, 4))
+
+    model = model_func()
+    learn = Learner(dls.to(def_device), model.to(def_device), loss_func=loss, metrics=[KLD, JSD])
+
+    if lr_find:
+        lr_suggestion = learn.lr_find()
+        plt.show()
+        plt.close()
+        lr = lr_suggestion.valley if hasattr(lr_suggestion, 'valley') else float(lr_suggestion)
+        print('lr_find selected', lr)
+
+    print('lr in training is', lr)
+    learn.fit_one_cycle(n_epoch, lr)
+
+    if save is not None:
+        learn.save(save)
+
+    pred, target = learn.get_preds()
+    pred = F.softmax(pred, dim=1).reshape(len(valid), -1)
+    target = target.reshape(len(valid), -1)
+
+    pred = pd.DataFrame(pred.detach().cpu().numpy(), index=valid.index, columns=target_col)
+    target = pd.DataFrame(target.detach().cpu().numpy(), index=valid.index, columns=target_col)
+    return target, pred
+
+# %% ../nbs/02_DNN.ipynb #a789a82c
+def predict_dl(df, feat_col, target_col, model_func, model_pth, A: int = 23, bs: int = 512):
+    "Predict a dataframe given a deep learning model saved by fastai." 
+    test_dset = GeneralDataset(df, feat_col, A=A)
+    test_dl = DataLoader(test_dset, batch_size=bs)
+
+    model = model_func()
+    infer_dls = DataLoaders.from_dsets(test_dset, test_dset, bs=bs, num_workers=min(fc.defaults.cpus, 4))
+    learn = Learner(infer_dls.to(def_device), model.to(def_device), loss_func=CE)
+    learn.load(model_pth, weights_only=False)
+    learn.model.eval()
+
+    preds = []
+    for data in test_dl:
+        inputs = data.to(def_device)
+        pred = learn.model(inputs)
+        pred = F.softmax(pred, dim=1).reshape(len(pred), -1)
+        preds.append(pred.detach().cpu().numpy())
+
+    preds = np.concatenate(preds)
+    return pd.DataFrame(preds, index=df.index, columns=target_col)
+
+# %% ../nbs/02_DNN.ipynb #b266abbf
+def train_dl_cv(df, feat_col, target_col, splits, model_func, A: int = 23, save: str | None = None, **kwargs):
+    "Cross-validation training loop for deep learning models." 
+    oof_frames = []
+    for fold, split in enumerate(splits):
+        print(f'------fold{fold}------')
+        fname = f'{save}_fold{fold}' if save is not None else None
+        _, pred = train_dl(df, feat_col, target_col, split, model_func, A=A, save=fname, **kwargs)
+        pred['nfold'] = fold
+        oof_frames.append(pred)
+    return pd.concat(oof_frames).sort_index()

kmodel-0.0.1/kmodel/train.py

@@ -0,0 +1,205 @@
+"""Generic helpers for tabular multi-output training and scoring, with runnable examples built from a seaborn dataset."""
+
+# AUTOGENERATED! DO NOT EDIT! File to edit: ../nbs/01_ML.ipynb.
+
+# %% auto #0
+__all__ = ['get_score_jsd', 'get_score_kld', 'get_score_ce', 'get_splits', 'split_data', 'train_ml', 'train_ml_cv',
+           'post_process', 'post_process_oof', 'get_score', 'js_divergence_flat', 'kl_divergence_flat', 'calculate_ce',
+           'predict_ml']
+
+# %% ../nbs/01_ML.ipynb #c4049499
+from collections.abc import Callable, Sequence
+from functools import partial
+from pathlib import Path
+
+import numpy as np
+import pandas as pd
+from joblib import dump, load
+from sklearn import set_config
+from sklearn.linear_model import LinearRegression
+from sklearn.model_selection import GroupKFold, StratifiedGroupKFold, StratifiedKFold
+
+set_config(transform_output="pandas")
+
+# %% ../nbs/01_ML.ipynb #a38f6af8
+def get_splits(
+    df: pd.DataFrame,
+    stratified: str | None = None,  # col used for stratified sampling
+    group: str | None = None,       # col used to keep grouped rows together
+    nfold: int = 5,
+    seed: int = 123,
+) -> list[tuple[np.ndarray, np.ndarray]]:
+    "Split samples in a dataframe with stratified, grouped, or stratified-grouped K-fold logic."
+
+    def _log(colname: str) -> None:
+        print(kf)
+        split = splits[0]
+        print(f"# {colname} in train set: {df.iloc[split[0]][colname].nunique()}")
+        print(f"# {colname} in test set: {df.iloc[split[1]][colname].nunique()}")
+
+    splits: list[tuple[np.ndarray, np.ndarray]] = []
+    if stratified is not None and group is None:
+        kf = StratifiedKFold(nfold, shuffle=True, random_state=seed)
+        for split in kf.split(df.index, df[stratified]):
+            splits.append(split)
+        _log(stratified)
+    elif group is not None and stratified is None:
+        kf = GroupKFold(nfold)
+        for split in kf.split(df.index, groups=df[group]):
+            splits.append(split)
+        _log(group)
+    elif stratified is not None and group is not None:
+        kf = StratifiedGroupKFold(nfold, shuffle=True, random_state=seed)
+        for split in kf.split(df.index, groups=df[group], y=df[stratified]):
+            splits.append(split)
+        _log(stratified)
+    else:
+        raise ValueError("Either 'stratified' or 'group' argument must be provided.")
+    return splits
+
+# %% ../nbs/01_ML.ipynb #efeeef7f
+def split_data(
+    df: pd.DataFrame,                 # dataframe of values
+    feat_col: Sequence[str],          # feature columns
+    target_col: Sequence[str],        # target columns
+    split: tuple[np.ndarray, np.ndarray],
+) -> tuple[pd.DataFrame, pd.DataFrame, pd.DataFrame, pd.DataFrame]:
+    "Given a split tuple, return X_train, y_train, X_test, and y_test." 
+    X_train = df.iloc[split[0]][list(feat_col)]
+    y_train = df.iloc[split[0]][list(target_col)]
+    X_test = df.iloc[split[1]][list(feat_col)]
+    y_test = df.iloc[split[1]][list(target_col)]
+    return X_train, y_train, X_test, y_test
+
+# %% ../nbs/01_ML.ipynb #9711d171
+def train_ml(
+    df: pd.DataFrame,                 # dataframe of values
+    feat_col: Sequence[str],          # feature columns
+    target_col: Sequence[str],        # target columns
+    split: tuple[np.ndarray, np.ndarray],
+    model,                            # sklearn model instance
+    save: str | Path | None = None,   # output path for joblib model
+    params: dict | None = None,       # kwargs forwarded to model.fit
+) -> tuple[pd.DataFrame, pd.DataFrame]:
+    "Fit and predict with a sklearn model, returning validation targets and predictions." 
+    X_train, y_train, X_test, y_test = split_data(df, feat_col, target_col, split)
+    model.fit(X_train, y_train, **(params or {}))
+    if save is not None:
+        dump(model, save)
+    y_pred = model.predict(X_test)
+    pred = pd.DataFrame(y_pred, index=y_test.index, columns=y_test.columns)
+    return y_test, pred
+
+# %% ../nbs/01_ML.ipynb #e2d3abaf
+def train_ml_cv(
+    df: pd.DataFrame,                             # dataframe of values
+    feat_col: Sequence[str],                      # feature columns
+    target_col: Sequence[str],                    # target columns
+    splits: Sequence[tuple[np.ndarray, np.ndarray]],
+    model,                                        # sklearn model instance
+    save: str | None = None,                      # model name prefix for saved folds
+    params: dict | None = None,                   # kwargs forwarded to model.fit
+) -> pd.DataFrame:
+    "Run cross-validation through the given splits." 
+    oof_frames: list[pd.DataFrame] = []
+    for fold, split in enumerate(splits):
+        save_path = f"models/{save}_{fold}.joblib" if save is not None else None
+        _, pred = train_ml(df, feat_col, target_col, split, model, save=save_path, params=params)
+        pred["nfold"] = fold
+        oof_frames.append(pred)
+    return pd.concat(oof_frames).sort_index()
+
+# %% ../nbs/01_ML.ipynb #f52b3804
+def _normalize_vector(values: pd.Series | np.ndarray, epsilon: float = 1e-8) -> np.ndarray:
+    "Normalize a probability-like vector after clipping small or negative values." 
+    array = np.asarray(values, dtype=float)
+    array = np.clip(array, epsilon, None)
+    return array / array.sum()
+
+# %% ../nbs/01_ML.ipynb #63757b94
+def post_process(pred_like: pd.DataFrame | pd.Series, epsilon: float = 1e-8):
+    "Clip negatives and renormalize probability-like predictions." 
+    if isinstance(pred_like, pd.Series):
+        values = _normalize_vector(pred_like, epsilon=epsilon)
+        return pd.Series(values, index=pred_like.index)
+
+    pred = pred_like.copy()
+    normalized = np.vstack([
+        _normalize_vector(row.to_numpy(dtype=float), epsilon=epsilon) for _, row in pred.iterrows()
+    ])
+    pred.loc[:, :] = normalized
+    return pred
+
+# %% ../nbs/01_ML.ipynb #f631adc4
+def post_process_oof(oof_ml: pd.DataFrame, target_col: Sequence[str]) -> pd.DataFrame:
+    "Post-process prediction columns in an out-of-fold dataframe." 
+    oof = oof_ml.copy()
+    oof[list(target_col)] = post_process(oof[list(target_col)])
+    return oof
+
+# %% ../nbs/01_ML.ipynb #64338f24
+def get_score(
+    target: pd.DataFrame,
+    pred: pd.DataFrame,
+    func: Callable[[pd.Series | np.ndarray, pd.Series | np.ndarray], float],
+) -> pd.Series:
+    "Apply a row-wise score function to aligned target and prediction frames." 
+    distance = [func(target.loc[i], pred.loc[i, target.columns]) for i in target.index]
+    return pd.Series(distance, index=target.index)
+
+# %% ../nbs/01_ML.ipynb #14487edb
+def js_divergence_flat(
+    target_series: pd.Series | np.ndarray,
+    pred_series: pd.Series | np.ndarray,
+    epsilon: float = 1e-8,
+) -> float:
+    "Compute Jensen-Shannon divergence between two flattened probability vectors." 
+    p = _normalize_vector(target_series, epsilon=epsilon)
+    q = _normalize_vector(pred_series, epsilon=epsilon)
+    m = 0.5 * (p + q)
+    return float(0.5 * np.sum(p * np.log(p / m)) + 0.5 * np.sum(q * np.log(q / m)))
+
+# %% ../nbs/01_ML.ipynb #06480577
+get_score_jsd = partial(get_score, func=js_divergence_flat)
+
+# %% ../nbs/01_ML.ipynb #9ec84ddd
+def kl_divergence_flat(
+    target_series: pd.Series | np.ndarray,
+    pred_series: pd.Series | np.ndarray,
+    epsilon: float = 1e-8,
+) -> float:
+    "Compute KL divergence between two flattened probability vectors." 
+    p = _normalize_vector(target_series, epsilon=epsilon)
+    q = _normalize_vector(pred_series, epsilon=epsilon)
+    return float(np.sum(p * np.log(p / q)))
+
+# %% ../nbs/01_ML.ipynb #5cacf75a
+get_score_kld = partial(get_score, func=kl_divergence_flat)
+
+# %% ../nbs/01_ML.ipynb #e1fe96d8
+def calculate_ce(
+    target_series: pd.Series | np.ndarray,
+    pred_series: pd.Series | np.ndarray,
+    epsilon: float = 1e-8,
+) -> float:
+    "Compute cross-entropy between two flattened probability vectors." 
+    p = _normalize_vector(target_series, epsilon=epsilon)
+    q = _normalize_vector(pred_series, epsilon=epsilon)
+    return float(-(p * np.log(q)).sum())
+
+# %% ../nbs/01_ML.ipynb #e173c2aa
+get_score_ce = partial(get_score, func=calculate_ce)
+
+# %% ../nbs/01_ML.ipynb #60892cc6
+def predict_ml(
+    df: pd.DataFrame,                     # dataframe with features
+    feat_col: Sequence[str],              # feature columns
+    target_col: Sequence[str] | None = None,
+    model_pth: str | Path = "model.joblib",
+) -> pd.DataFrame:
+    "Predict from a saved sklearn model." 
+    test = df[list(feat_col)]
+    model = load(model_pth)
+    pred = model.predict(test)
+    columns = list(target_col) if target_col is not None else list(range(pred.shape[1]))
+    return pd.DataFrame(pred, index=df.index, columns=columns)

kmodel-0.0.1/kmodel.egg-info/PKG-INFO

@@ -0,0 +1,89 @@
+Metadata-Version: 2.4
+Name: kmodel
+Version: 0.0.1
+Summary: A fresh nbdev project for kmodel.
+Author-email: keke <lcai888666@gmail.com>
+License: Apache-2.0
+Project-URL: Repository, https://github.com/sky1ove/kmodel
+Project-URL: Documentation, https://sky1ove.github.io/kmodel/
+Keywords: nbdev
+Classifier: Programming Language :: Python :: 3
+Classifier: Programming Language :: Python :: 3 :: Only
+Requires-Python: >=3.10
+Description-Content-Type: text/markdown
+License-File: LICENSE
+Requires-Dist: fastai>=2.8.7
+Requires-Dist: joblib>=1.5.3
+Requires-Dist: numpy>=2.2.6
+Requires-Dist: pandas>=2.3.3
+Requires-Dist: scikit-learn>=1.7.2
+Requires-Dist: torch>=2.11.0
+Dynamic: license-file
+
+# kmodel
+
+
+<!-- WARNING: THIS FILE WAS AUTOGENERATED! DO NOT EDIT! -->
+
+This file will become your README and also the index of your
+documentation.
+
+## Developer Guide
+
+If you are new to using `nbdev` here are some useful pointers to get you
+started.
+
+### Install kmodel in Development mode
+
+``` sh
+# make sure kmodel package is installed in development mode
+$ pip install -e .
+
+# make changes under nbs/ directory
+# ...
+
+# compile to have changes apply to kmodel
+$ nbdev_prepare
+```
+
+## Usage
+
+### Installation
+
+Install latest from the GitHub
+[repository](https://github.com/sky1ove/kmodel):
+
+``` sh
+$ pip install git+https://github.com/sky1ove/kmodel.git
+```
+
+or from [conda](https://anaconda.org/sky1ove/kmodel)
+
+``` sh
+$ conda install -c sky1ove kmodel
+```
+
+or from [pypi](https://pypi.org/project/kmodel/)
+
+``` sh
+$ pip install kmodel
+```
+
+### Documentation
+
+Documentation can be found hosted on this GitHub
+[repository](https://github.com/sky1ove/kmodel)’s
+[pages](https://sky1ove.github.io/kmodel/). Additionally you can find
+package manager specific guidelines on
+[conda](https://anaconda.org/sky1ove/kmodel) and
+[pypi](https://pypi.org/project/kmodel/) respectively.
+
+## How to use
+
+Fill me in please! Don’t forget code examples:
+
+``` python
+1+1
+```
+
+    2

kmodel-0.0.1/kmodel.egg-info/SOURCES.txt

@@ -0,0 +1,15 @@
+LICENSE
+MANIFEST.in
+README.md
+pyproject.toml
+kmodel/__init__.py
+kmodel/_modidx.py
+kmodel/core.py
+kmodel/dnn.py
+kmodel/train.py
+kmodel.egg-info/PKG-INFO
+kmodel.egg-info/SOURCES.txt
+kmodel.egg-info/dependency_links.txt
+kmodel.egg-info/entry_points.txt
+kmodel.egg-info/requires.txt
+kmodel.egg-info/top_level.txt

kmodel-0.0.1/kmodel.egg-info/dependency_links.txt

@@ -0,0 +1 @@
+

kmodel-0.0.1/kmodel.egg-info/entry_points.txt

@@ -0,0 +1,2 @@
+[nbdev]
+kmodel = kmodel._modidx:d

kmodel-0.0.1/kmodel.egg-info/requires.txt

@@ -0,0 +1,6 @@
+fastai>=2.8.7
+joblib>=1.5.3
+numpy>=2.2.6
+pandas>=2.3.3
+scikit-learn>=1.7.2
+torch>=2.11.0

kmodel-0.0.1/kmodel.egg-info/top_level.txt

@@ -0,0 +1 @@
+kmodel

kmodel-0.0.1/pyproject.toml

@@ -0,0 +1,50 @@
+[build-system]
+requires = ["setuptools>=64"]
+build-backend = "setuptools.build_meta"
+
+[project]
+name = "kmodel"
+dynamic = ["version"]
+description = "A fresh nbdev project for kmodel."
+readme = "README.md"
+requires-python = ">=3.10"
+license = {text = "Apache-2.0"}
+authors = [{name = "keke", email = "lcai888666@gmail.com"}]
+keywords = ['nbdev']
+classifiers = [
+    "Programming Language :: Python :: 3",
+    "Programming Language :: Python :: 3 :: Only",
+]
+dependencies = [
+    "fastai>=2.8.7",
+    "joblib>=1.5.3",
+    "numpy>=2.2.6",
+    "pandas>=2.3.3",
+    "scikit-learn>=1.7.2",
+    "torch>=2.11.0",
+]
+
+[project.urls]
+Repository = "https://github.com/sky1ove/kmodel"
+Documentation = "https://sky1ove.github.io/kmodel/"
+
+[project.entry-points.nbdev]
+kmodel = "kmodel._modidx:d"
+
+[tool.setuptools.dynamic]
+version = {attr = "kmodel.__version__"}
+
+[tool.setuptools.packages.find]
+include = ["kmodel"]
+
+[tool.nbdev]
+title = "kmodel"
+lib_path = "kmodel"
+
+[dependency-groups]
+dev = [
+    "ipykernel>=7.2.0",
+    "nbclient>=0.10.4",
+    "nbformat>=5.10.4",
+    "seaborn>=0.13.2",
+]

kmodel-0.0.1/setup.cfg

@@ -0,0 +1,4 @@
+[egg_info]
+tag_build = 
+tag_date = 0
+