Pinebioml

description:
1. check pipenv and readme

description:
1. sklearn pipeline support
a. some api changes: select, imputer
2. Normalizer reconstructed
3. Adding example:
a. example_PineLine: the example of using sklearn pipeline.
b. example_UsingExistingModel: the example of using existing models.
4. Optuna tuning target changes
from: score
to: test_score + 0.2 * (test_score - train_score)
5. document update.
bug:
1. bagging:
- contradicts to log domain in selection method (the negative output of pca)
- the variance of feature will change during bagging (variance matters)
2. Lasso logistic regression in grid search need to to modified. In temporary using binary search.
3. Issue about installation of pandas lacking of Microsoft Visual C++ 14.0:
https://learn.microsoft.com/zh-tw/cpp/windows/latest-supported-vc-redist?view=msvc-170
On Going:
1. Testing and Deploy (Done)
2. A reliable tutorial. (including mac)
3. One click everything
ToDo:
1. report and diagnose
2. Document reference
3. using pretty, beautiful, good-looking, precise packages:
a. pca
b. The only OPLS da reliable(compare to others), alive, python implement
https://github.com/Omicometrics/pypls?tab=readme-ov-file
n. add parameter dict(json or yaml-like)
n. interactive interface or GUI (maybe nicegui/ plotly)

description:
1. tuner structure adjustment:
a. evaluate now implement under basic tuner.
b. tuner now have 3 random seeds: kernel_seed for model, valid_seed for cv, optuna_seed for optuna.
c. add comparison between optuna result and default parameter setting.
2. tuner regression was finished but not sufficiently tested.
3. tuner name changed: SVC_tuner->SVM_tuner
4. tuner add xgboost and lightgbm.
5. defualt tuner parameters changes
6. document updates
bug:
1. bagging:
- contradicts to log domain in selection method (the negative output of pca)
- the variance of feature will change during bagging (variance matters)
2. Lasso logistic regression in grid search need to to modified. In temporary using binary search.
3. Issue about installation of pandas lacking of Microsoft Visual C++ 14.0
On Going:
1. Testing and Deploy
2. A reliable tutorial. (including mac)
ToDo:
1. report and diagnose
2. Document reference
3. using pretty, beautiful, good-looking, precise packages:
a. pca
b. The only OPLS da reliable(compare to others), alive, python implement
https://github.com/Omicometrics/pypls?tab=readme-ov-file
4. One click everything
n. add parameter dict(json or yaml-like)
n. interactive interface or GUI (maybe nicegui/ plotly)

description:
1. change the order of tuner data pipeline
2. defualt tuner parameters changes
3. add statsmodels summary (R style) into elasticnet tuner
4. the name of ElasticNet turner was changed to ElasticLogit
5. add fit and predict api to tuner (for potential usage of sklearn pipeline)

description:
1. defualt tuner target changes to mcc
2. defualt tuner parameters changes

description:
1. bug fixed:
a. IO.read_file can read tsv now.
2. report utils:
a. data_overview:
I. Abandon boxplot and pairplot.
II. PCA use 4 component.
III. add UMAP, PLS
ToDo: add pacmap and hysterical clustering
3. auto modeling:
a. optuna model tuner for ElasticNet.
b. stablity of svm tuner
4. preprocessing.utils
a. feature_extension: extend input features by a*b, arctan a/b, pca component.
5. tutorial updating
6. ...etc (sorry I forget what I changed...)

On going:
1. auto modeling:
a. Performance oriented modeling.
b. optuna: auto tuner for all models
2. Diagnose and report

bug:
1. bagging:
- contradicts to log domain in selection method (the negative output of pca)
- the variance of feature will change during bagging (variance matters)
2. Lasso logistic regression in grid search need to to modified. In temporary using binary search.

to do:
0. Tutorial for mac
1. Document and reference
2. using pretty, beautiful, good-looking, precise packages:
a. pca
b. The only OPLS da reliable(compare to others), alive, python implement
https://github.com/Omicometrics/pypls?tab=readme-ov-file
3. One click everything
n. add parameter dict(json or yaml-like)
n. interactive interface or GUI (maybe nicegui/ plotly)