mindstrong 1.0

安装

安装软件包的最简单方法是通过easy_install或pip：

$ pip install mindstrong_biomarker_modelfit

这也应该考虑依赖关系（numpy、scipy、pandas和sklearn）。

用法

该项目包括模拟数字生物标志物和目标测量数据。要使模型适合这些示例数据：

import numpy as np
import pandas as pd
import os
from mindstrong import mindstrong_modelfit as mindstrong

target_file = mindstrong.get_example_data('example_targets.csv')
feature_file = mindstrong.get_example_data('example_features.csv')
target_colname = 'target1'

# Load target data
target_df = pd.read_csv(target_file)
target_df.set_index('device_id', inplace=True)

# Load Feature Data
feature_df = pd.read_csv(feature_file).set_index(['device_id', 'targetDOY'])

# Cross Validated supervised kernel PCA model-fitting
cvdf, best_model = mindstrong.calculateCrossValidatedCorrelation(target_df,
                                                                 feature_df,
                                                                 target_colname,
                                                                 fold_type='n',
                                                                 n_folds=5,
                                                                 kernel_training='linear',
                                                                 kernel_training_param=1,
                                                                 kernel_target='linear',
                                                                 kernel_target_param=1,
                                                                 regularization=0.1)

# Print the final results
print(best_model)

版权和许可证

版权所有（c）2018，Mindstrong Health。GNU Affero通用公共许可证。