def main():
"""Ensemble all models inside the experiments folder"""
# we assume all the experiments are saved
# in the experiments folder
path = Path('experiments')
# get a list of all experiments name
experiment_list = os.listdir(path)
assert len(experiment_list) > 1, \
'there is not enough experiments to ensemble'
predictions = []
# for every experiment
for experiment in experiment_list:
# create a path to the valid prediction file
path_to_pred = path.joinpath(experiment, 'prediction', 'valid.csv')
if not os.path.exists(path_to_pred):
continue
# if this file exists, we read it and
# set the experiment column to the name of this experiment
pred_exp = load_data.read_csv(path_to_pred)
pred_exp = pred_exp.assign(experiment=experiment)
predictions.append(pred_exp)
# concat all the predictions
predictions = pd.concat(predictions)
# create the target by dropping all duplicates
target = predictions.drop_duplicates(subset=['period', 'timedelta'])
target.reset_index(drop=True, inplace=True)
target.drop(columns=default.yhat, inplace=True)
# ensemble
predictions_ensemble = ensemble(predictions)
target_ensemble = target.merge(predictions_ensemble,
on=['period', 'timedelta'],
how='left')
# check there is non nan values
assert target_ensemble[default.yhat].isna().sum().sum() == 0
# compute the metrics
ensemble_metrics = compute_metrics(target_ensemble)
experiment_list = list(predictions['experiment'].unique())
ensemble_metrics['experiment'] = '__'.join(experiment_list)
ensemble_metrics['n_model'] = len(experiment_list)
results = pd.DataFrame([ensemble_metrics])
# print scores
print(results.head())
# save the ensemble results in a CSV file
results.to_csv(path / 'ensemble_summary.csv', index=False)
def main():
"""
This function will save the solar wind data
as a Feather file
"""
# read the main config file
config = load_data.read_config_file('./config/config.yml')
# get the path to the CSV File
directories = config['directories']
raw_path = Path(directories['raw'])
interim_path = Path(directories['interim'])
interim_path.mkdir(exist_ok=True, parents=True)
logging.info('reading solar wind data..')
# reading CSV file
solar_wind = load_data.read_csv(raw_path / 'solar_wind.csv')
logging.info('saving to feather..')
# saving as feather file
solar_wind.to_feather(interim_path / 'solar_wind.feather')
from load_data import read_csv
path = "data/sample_submission.csv"
if __name__ == "__main__":
""" Main script to run analysis"""
df = read_csv(path)
def main(use_sample: bool = False, n_jobs: int = 1):
"""
This function will apply all the steps in order to create
a dataset ready to train models.
The following steps:
- read the data
- compute the solar wind features
- compute sattelite positions features
- take the log of smoothed_ssn values
- create the target for the actual time t and t + 1 hour
- merge all dataset into a single one
- save the dataset for future modeling
# Params
use_sample: `bool`, optional(defualt=False)
Whether or not to use the sample dataset
n_jobs: `in`, optional(defualt=1)
The number of jobs to run in parallel
"""
logging.info(f'use_sample={use_sample}, n_jobs={n_jobs}')
logging.info('reading config file')
config = load_data.read_config_file('./config/config.yml')
# directories
directories = config['directories']
raw_path = Path(directories['raw'])
interim_path = Path(directories['interim'])
processed_path = Path(directories['processed'])
processed_path.mkdir(exist_ok=True, parents=True)
# reading gt data
solar_wind_file = ('sample_solar_wind.feather'
if use_sample else 'solar_wind.feather')
logging.info('reading training data')
dst_labels = load_data.read_csv(raw_path / 'dst_labels.csv')
solar_wind = load_data.read_feather(interim_path / solar_wind_file)
sunspots = load_data.read_csv(raw_path / 'sunspots.csv')
stl_pos = load_data.read_csv(raw_path / 'satellite_positions.csv')
logging.info('preprocessing solar wing')
# preprocessing solar wind
# setting timedelta as index
solar_wind.set_index('timedelta', inplace=True)
# preprocessing solar wind time series
solar_wind = solar_wind_preprocessing(solar_wind)
logging.info('computing features')
start = time.time()
# computing solar wind features
data = split_into_period(solar_wind,
features=default.init_features,
n_jobs=n_jobs)
elapsed_time = (time.time() - start) / 60
logging.info(f'elapsed time {elapsed_time:.4f}')
logging.info('merging other datasets')
# create target
target = create_target(dst_labels)
# preprocessing sattelite positions
stl_pos = stl_preprocessing(stl_pos)
# taking the log of smoothed_ssn values
sunspots['smoothed_ssn'] = np.log(sunspots['smoothed_ssn'])
# merging dataframes to the main dataframe
data = merge_daily(data, stl_pos)
data = merge_daily(data, sunspots)
# merging target dataframe to the main dataframe
data = data.merge(target, how='left', on=['period', 'timedelta'])
# droping last values where there is not available data
data.dropna(subset=['t0', 't1'], inplace=True)
# reset index
data.reset_index(inplace=True, drop=True)
logging.info('saving')
output_filename = 'fe' if not use_sample else 'fe_sample'
# saving to feather format
data.to_feather(processed_path / f'{output_filename}.feather')
import import_path
import numpy as np
import pandas as pd
from models.oracle import Oracle
from models.surrogate_teacher import Surrogate
from models.omniscient_teacher import Omniscient
from models.random_teacher import Random
from models.without_teacher import Without_teacher
from utils import predict, predict_by_W, rmse_W, write_np2csv, rmse_w, make_random_mask, predict_wj
from load_data import read_W, read_csv, split_data
from tqdm import tqdm
from sklearn.metrics import roc_auc_score
import logging
import datetime
# %%
df = read_csv('output/wine-quality-pm1.csv', header=0)
train_X, test_X, train_y, test_y = split_data(df, True)
eta, lambd, alpha = 1, 2, 0.01
training_epochs, loops = 10, 10
J = 10
# 提示する教材合計数
textbook = 500
# 推定に使う教材数
test_textbook_list = [100]
# 推定間に提示する教材数
between_textbook_list = [1]
# 組
k = 1
lambds = [1, 2, 3, 4, 5]
for lambd in lambds:
import numpy as np
import pandas as pd
from models.oracle import Oracle
from models.surrogate_teacher import Surrogate
from models.omniscient_teacher import Omniscient
from models.random_teacher import Random
from models.without_teacher import Without_teacher
from utils import predict, predict_by_W, rmse_W, write_np2csv, rmse_w, make_random_mask, predict_wj
from load_data import read_W, read_csv, split_data
from tqdm import tqdm
from sklearn.metrics import roc_auc_score
import logging
import datetime
# %%
df = read_csv('output/weebil_vespula_pm1.csv', header=0)
train_X, test_X, train_y, test_y = split_data(df, False)
eta, lambd, alpha = 1, 2, 0.01
training_epochs, loops = 10, 10
J = 10
# 提示する教材合計数
textbook = 500
# 推定に使う教材数
test_textbook_list = [100]
# 推定間に提示する教材数
between_textbook_list = [1]
# 組
k = 1
lambds = [1, 2, 3, 4, 5]
for lambd in lambds:
from load_data import read_csv
from svd import SVD
from sklearn.cluster import MiniBatchKMeans
import pandas as pd
if __name__ == '__main__':
rec_data = read_csv('../input/user_item_cnt.csv')
rec = SVD(rec_data)
rec.fit()
score = rec.get_score()
tmp = [dict(v, user_id=user_id) for user_id, aaa in score.items() for v in aaa]
df = pd.DataFrame(tmp)
df.head()
df.to_csv('svd2.csv', index=False)
"""
model = MiniBatchKMeans(n_clusters=100, random_state=0)
model.fit(rec.user_matrix)
pred = model.predict(rec.user_matrix)
users = [rec_data.map_idx2user[i] for i in range(len(rec_data.map_idx2user))]
max(users)
len(rec_data.map_idx2user)
df = pd.DataFrame({'user_id': users, 'cluster': pred})
df.to_csv('cluster.csv', index=False)
"""
# Optional check for unexpected values
if not np.isfinite(prediction_at_t0):
prediction_at_t0 = -12
if not np.isfinite(prediction_at_t1):
prediction_at_t1 = -12
return prediction_at_t0, prediction_at_t1
if __name__ == '__main__':
# We use this code for testing
import load_data
import time
raw_path = Path('data/raw/')
interim_path = Path('data/interim')
dst_labels = load_data.read_csv(raw_path / 'dst_labels.csv')
solar_wind = load_data.read_feather(interim_path / 'solar_wind.feather')
sunspots = load_data.read_csv(raw_path / 'sunspots.csv')
stl_pos = load_data.read_csv(raw_path / 'satellite_positions.csv')
date = pd.to_timedelta(7, unit='d')
# date = pd.to_timedelta('111 days 04:00:00')
one_minute = pd.to_timedelta("1 minute")
seven_days = pd.to_timedelta("7 days")
solar_wind = solar_wind[solar_wind['period'] == 'train_a']
sunspots = sunspots[sunspots['period'] == 'train_a']
stl_pos = stl_pos[stl_pos['period'] == 'train_a']
solar_wind.set_index(['timedelta'], inplace=True)
stl_pos.set_index(['timedelta'], inplace=True)
sunspots.set_index(['timedelta'], inplace=True)
t_minus_7 = date - seven_days
c[j][0] += (math.pow(time[i], j) * price[i])
matrix_c = np.matrix(c)
# Now Calc mean
mean_matrix = matrix_a * matrix_S * matrix_c * beta
mean = mean_matrix.item(0)
# Now Calc 𝜑(x)
d = [[0] for _ in range(M + 1)]
for i in range(M + 1):
d[i][0] = math.pow(test_val, i)
matrix_d = np.matrix(d)
# Now Calc variance
variance = math.sqrt((matrix_a * matrix_S * matrix_d)[0][0] + (1 / beta))
# Now show results
print("Predicted Val : {:.4f}".format(mean))
print("Actual Val : {:.4f}".format(price[-1]))
print("Range Predction : [{:.4f}, {:.4f}]".format(mean - 3 * variance,
mean + 3 * variance))
print("Absolute Error : {:.4f}".format(abs(price[-1] - mean)))
print("Relative Error : {:.4f}%".format(
abs(price[-1] - mean) / price[-1] * 100))
if __name__ == "__main__":
csv_path = "./data/"
files = load_data.find_csv(csv_path)
for f in files:
print(("-" * 5 + f.split(".")[0] + " Summary" + "-" * 20)[:40])
time, price = load_data.read_csv(csv_path, f)
baysian_regression_predict(time, price, time[-1])
print("-" * 40, end="\n\n\n")
