def run_titanic():
df_titanic = pd.read_csv(
'https://raw.githubusercontent.com/datasciencedojo/datasets/master/titanic.csv'
)
df_titanic = df_titanic[[
'Survived', 'Pclass', 'Sex', 'Age', 'SibSp', 'Parch', 'Fare'
]]
# df_titanic['Pclass'] = df_titanic['Pclass'].map(lambda x: str(x)+' class')
df_titanic['Pclass'] = np.array(
['zero', 'class one', 'class two',
'class three'])[df_titanic['Pclass']]
df_titanic['Survived'] = np.array([True, False])[df_titanic['Survived']]
autoregression.compare_predictions(
df_titanic,
'survived',
# percent_data=1,
corr_matrix=True,
# scatter_matrix=True, #doesn't work IF categorical values of 4 groups in X.
bootstrap_coefs=True,
partial_dep=True,
plot_alphas=True,
plot_predicted_vs_actuals_flag=True,
plot_coefs_flag=True,
feature_importances=True,
actual_vs_predicted=True,
plot_predicteds_vs_actuals=True,
residuals=True,
univariates=True,
compare_models=True,
ROC=True,
show_plots=True,
)
def run_iris():
iris_df = pd.read_csv('https://raw.githubusercontent.com/mwaskom/seaborn-data/master/iris.csv')
# autoregression.compare_predictions(iris_df,'sepal_length')
iris_df['foods'] = np.random.choice(['hot dogs', 'bacon', 'sweets', np.NaN, np.inf], iris_df.shape[0], )
autoregression.compare_predictions(iris_df,'sepal_length', percent_data=1,
# corr_matrix=True,
scatter_matrix=True,
bootstrap_coefs=True,
partial_dep=True,
plot_predicted_vs_actuals_flag=True,
plot_coefs_flag=True,
feature_importances=True,
actual_vs_predicted=True,
plot_predicteds_vs_actuals=True,
residuals=True,
univariates=True,
compare_models=True,
ROC=True
)
def main():
posts = load_posts()
classrooms = load_classrooms()
posts['year_month'] = pd.to_datetime(posts['date']).map(lambda dt: dt.replace(day=1))
classrooms_merged = pd.read_csv('../data_january/classrooms_merged_non_leak.csv')
now = declare_now()
six_months = timedelta(days=365//2)
# one_year = timedelta(days=365)
last_month_posts_total = get_posts_total_between_with_zeros(posts, now, six_months)
full_y, y = make_y_was_any_posts_between(posts, classrooms, now, six_months)
# print(find_y_hearts(posts, now))
# print(find_y_hearts(posts[posts['classroom_id']==1 | posts['classroom_id']==3], now))
# y_hearts = get_all_y_hearts(posts, now)
classrooms_merged_before_now = make_classrooms_merged_before(classrooms_merged, now)
classrooms_merged_before_three_months_ago = make_classrooms_merged_before(classrooms_merged, now-timedelta(days=90))
classrooms_merged_before_now_dropped = drop_collumns(classrooms_merged_before_now)
classrooms_merged_before_three_months_ago = drop_collumns(classrooms_merged_before_three_months_ago)
classrooms_sum_before_now = make_classrooms_merged_before(classrooms_merged, now)
classrooms_sum_before_now_dropped = make_classrooms_merged_before(classrooms_merged_before_now_dropped, now)
classrooms_sum_before_three_months_ago = make_classrooms_merged_before(classrooms_merged_before_three_months_ago, now)
check_merge_lengths(classrooms, classrooms_merged, posts, classrooms_merged_before_now_dropped, classrooms_sum_before_now, y, full_y)
classrooms_sum_before_three_months_ago = make_classrooms_sum_(classrooms_merged_before_three_months_ago, full_y)
classrooms_sum_before_now = make_classrooms_sum_(classrooms_merged, full_y)
# classrooms_sum_hearts_before_three_months_ago = merged_to_sum_before(classrooms_merged, now-timedelta(days=90), y_hearts)
# classrooms_sum_hearts_before_three_months_ago = make_classrooms_sum_hearts_before_now(classrooms_merged_before_three_months_ago)
# classrooms_sum_hearts_before_now = make_classrooms_sum_hearts_before_now(classrooms_merged)
# names3, results3, models3, pipeline3, df_X3 = autoregression.compare_predictions(classrooms_sum_hearts_before_three_months_ago, 'will_post_next_semester')
names2, results2, models2, pipeline2, df_X2 = autoregression.compare_predictions(classrooms_sum_before_three_months_ago, 'will_post_next_semester', corr_matrix=True,
scatter_matrix=False, bootstrap_coefs=False,
feature_importances=False,
partial_dep=False, actual_vs_predicted=False,
residuals=False, univariates=False, compare_models=False,
ROC=False)
output = open('random_forrest.pkl', 'wb')
pickle.dump(models2[3], output)
output.close()
df_X2.to_csv('df_X2.csv')
'Survived', 'Pclass', 'Sex', 'Age', 'SibSp', 'Parch', 'Fare'
]]
# df_titanic['Pclass'] = df_titanic['Pclass'].map(lambda x: str(x)+' class')
df_titanic['Pclass'] = np.array(
['zero', 'class one', 'class two', 'class three'])[df_titanic['Pclass']]
df_titanic['Survived'] = np.array([True, False])[df_titanic['Survived']]
names, results, fit_models, pipeline, df_X, y_hats, errors = autoregression.compare_predictions(
df_titanic,
'survived',
percent_data=1,
corr_matrix=True,
# scatter_matrix=True, #doesn't work IF categorical values of 4 groups in X.
bootstrap_coefs=True,
partial_dep=True,
plot_alphas=True,
plot_predicted_vs_actuals_flag=True,
plot_coefs_flag=True,
feature_importances=True,
actual_vs_predicted=True,
plot_predicteds_vs_actuals=True,
residuals=True,
univariates=True,
compare_models=True,
ROC=True,
)
# autoregression.compare_predictions(iris_df,'sepal_length',
# feature_importances=False
# )
from autoregression import clean_dataframe
import autoregression
import pandas as pd
import matplotlib
import numpy as np
np.random.seed(seed=99)
# matplotlib.interactive(True)
import matplotlib.pyplot as plt
from sklearn.datasets import load_wine
wine = load_wine()
df = pd.DataFrame(wine.data, columns=wine.feature_names)
df['foods'] = np.random.choice(['hot dogs', 'bacon', 'sweets', np.NaN, np.inf],
df.shape[0])
# df['wine_class'] = wine['target_names'][wine.target] # This has 3 classifications!!!
df['wine_class'] = (wine['target_names'][wine.target] == 'class_1')
autoregression.compare_predictions(df,
'wine_class',
corr_matrix=True,
scatter_matrix=True,
bootstrap_coefs=True,
partial_dep=True,
plot_predicted_vs_actuals=True,
plot_coefs_flag=True,
feature_importances=True,
actual_vs_predicted=True,
plot_predicteds_vs_actuals=True,
residuals=True,
univariates=True,
compare_models=True,
ROC=True)