markdown stringlengths 0 1.02M | code stringlengths 0 832k | output stringlengths 0 1.02M | license stringlengths 3 36 | path stringlengths 6 265 | repo_name stringlengths 6 127 |
|---|---|---|---|---|---|
EvaluationUsually, the model and its predictions is not sufficient. In the following we want to evaluate our classifiers. Let's start by computing their error. The sklearn.metrics package contains several errors such as* Mean squared error* Mean absolute error* Mean squared log error* Median absolute error | #computing the squared error of the first model
print("Mean squared error model 1: %.2f" % mean_squared_error(targetFeature1, targetFeature1_predict)) | Mean squared error model 1: 0.56
| MIT | Instruction4/Instruction4-RegressionSVM.ipynb | danikhani/ITDS-Instructions-WS20 |
We can also visualize the errors: | plt.scatter(targetFeature1_predict, (targetFeature1 - targetFeature1_predict) ** 2, color = "blue", s = 10,)
## plotting line to visualize zero error
plt.hlines(y = 0, xmin = 0, xmax = 15, linewidth = 2)
## plot title
plt.title("Squared errors Model 1")
## function to show plot
plt.show() | _____no_output_____ | MIT | Instruction4/Instruction4-RegressionSVM.ipynb | danikhani/ITDS-Instructions-WS20 |
Now it is your turn. Compute the mean squared error and visualize the squared errors. Play around using different error metrics. | #Your turn
print("Mean squared error model 2: %.2f" % mean_squared_error(targetFeature2,targetFeature2_predict))
print("Mean absolute error model 2: %.2f" % mean_absolute_error(targetFeature2,targetFeature2_predict))
plt.scatter(targetFeature2_predict, (targetFeature2 - targetFeature2_predict) ** 2, color = "blue",)
... | Mean squared error model 2: 8.89
Mean absolute error model 2: 2.32
| MIT | Instruction4/Instruction4-RegressionSVM.ipynb | danikhani/ITDS-Instructions-WS20 |
Handling multiple descriptive features at once - Multiple linear regressionIn most cases, we will have more than one descriptive feature . As an example we use an example data set of the scikit package. The dataset describes housing prices in Boston based on several attributes. Note, in this format the data is already... | from sklearn import datasets ## imports datasets from scikit-learn
df3 = datasets.load_boston()
#The sklearn package provides the data splitted into a set of descriptive features and a target feature.
#We can easily transform this format into the pandas data frame as used above.
descriptiveFeatures3 = pd.DataFrame(df3... | Descriptive features:
CRIM ZN INDUS CHAS NOX RM AGE DIS RAD TAX \
0 0.00632 18.0 2.31 0.0 0.538 6.575 65.2 4.0900 1.0 296.0
1 0.02731 0.0 7.07 0.0 0.469 6.421 78.9 4.9671 2.0 242.0
2 0.02729 0.0 7.07 0.0 0.469 7.185 61.1 4.9671 2.0 242.0
3 0.0... | MIT | Instruction4/Instruction4-RegressionSVM.ipynb | danikhani/ITDS-Instructions-WS20 |
To predict the housing price we will use a Multiple Linear Regression model. In Python this is very straightforward: we use the same function as for simple linear regression, but our set of descriptive features now contains more than one element (see above). | classifier = LinearRegression()
model3 = classifier.fit(descriptiveFeatures3,targetFeature3)
targetFeature3_predict = classifier.predict(descriptiveFeatures3)
print('Coefficients: \n', classifier.coef_)
print('Intercept: \n', classifier.intercept_)
print("Mean squared error: %.2f" % mean_squared_error(targetFeature3, ... | Coefficients:
[[-1.08011358e-01 4.64204584e-02 2.05586264e-02 2.68673382e+00
-1.77666112e+01 3.80986521e+00 6.92224640e-04 -1.47556685e+00
3.06049479e-01 -1.23345939e-02 -9.52747232e-01 9.31168327e-03
-5.24758378e-01]]
Intercept:
[36.45948839]
Mean squared error: 21.89
| MIT | Instruction4/Instruction4-RegressionSVM.ipynb | danikhani/ITDS-Instructions-WS20 |
As you can see above, we have a coefficient for each descriptive feature. Handling categorical descriptive featuresSo far we always encountered numerical dscriptive features, but data sets can also contain categorical attributes. The regression function can only handle numerical input. There are several ways to tranfo... | #example using pandas
df4 = pd.DataFrame({'A':['a','b','c'],'B':['c','b','a'] })
one_hot_pd = pd.get_dummies(df4)
one_hot_pd
#example using scikit
from sklearn.preprocessing import LabelEncoder, OneHotEncoder
#apply the one hot encoder
encoder = OneHotEncoder(categories='auto')
encoder.fit(df4)
df4_OneHot = encoder.tr... | Transformed by One-hot Encoding:
[[1. 0. 0. 0. 0. 1.]
[0. 1. 0. 0. 1. 0.]
[0. 0. 1. 1. 0. 0.]]
Replacing categories by numerical labels:
A B
0 0 2
1 1 1
2 2 0
| MIT | Instruction4/Instruction4-RegressionSVM.ipynb | danikhani/ITDS-Instructions-WS20 |
Now it is your turn. Perform linear regression using the data set given below. Don't forget to transform your categorical descriptive features. The rental price attribute represents the target variable. | from sklearn.preprocessing import LabelEncoder
df5 = pd.DataFrame({'Size':[500,550,620,630,665],'Floor':[4,7,9,5,8], 'Energy rating':['C', 'A', 'A', 'B', 'C'], 'Rental price': [320,380,400,390,385] })
#Your turn
# To transform the categorial feature
to_trannsform = df5[['Energy rating']]
encoder = LabelEncoder()
trans... | Coefficients:
[ 0.39008474 -0.54300185 -18.80539593]
Intercept:
166.068958800039
Mean squared error: 4.68
| MIT | Instruction4/Instruction4-RegressionSVM.ipynb | danikhani/ITDS-Instructions-WS20 |
Predicting a categorical target value - Logistic regression We might also encounter data sets where our target feature is categorical. Here we don't transform them into numerical values, but insetad we use a logistic regression function. Luckily, sklearn provides us with a suitable function that is similar to the line... | # Importing the dataset
iris = pd.read_csv('iris.csv')
print('First look at the data set: ')
print(iris.head())
#defining the descriptive and target features
descriptiveFeatures_iris = iris[['sepal_length']] #we only use the attribute 'sepal_length' in this example
targetFeature_iris = iris['species'] #we want to pre... | First look at the data set:
sepal_length sepal_width petal_length petal_width species
0 5.1 3.5 1.4 0.2 setosa
1 4.9 3.0 1.4 0.2 setosa
2 4.7 3.2 1.3 0.2 setosa
3 4.6 3.1 ... | MIT | Instruction4/Instruction4-RegressionSVM.ipynb | danikhani/ITDS-Instructions-WS20 |
Now it is your turn. In the example above we only used the first attribute as descriptive variable. Change the example such that all available attributes are used. | #Your turn
# Importing the dataset
iris2 = pd.read_csv('iris.csv')
print('First look at the data set: ')
print(iris2.head())
#defining the descriptive and target features
descriptiveFeatures_iris2 = iris[['sepal_length','sepal_width','petal_length','petal_width']]
targetFeature_iris2 = iris['species'] #we want to pr... | First look at the data set:
sepal_length sepal_width petal_length petal_width species
0 5.1 3.5 1.4 0.2 setosa
1 4.9 3.0 1.4 0.2 setosa
2 4.7 3.2 1.3 0.2 setosa
3 4.6 3.1 ... | MIT | Instruction4/Instruction4-RegressionSVM.ipynb | danikhani/ITDS-Instructions-WS20 |
Note, that the regression classifier (both logistic and non-logistic) can be tweaked using several parameters. This includes, but is not limited to, non-linear regression. Check out the documentation for details and feel free to play around! Support Vector Machines Aside from regression models, the sklearn package als... | from sklearn.svm import SVC
#define descriptive and target features as before
descriptiveFeatures_iris = iris[['sepal_length', 'sepal_width', 'petal_length', 'petal_width']]
targetFeature_iris = iris['species']
#this time, we train an SVM classifier
classifier = SVC(C=1, kernel='linear', gamma = 'auto')
classifier.fi... | _____no_output_____ | MIT | Instruction4/Instruction4-RegressionSVM.ipynb | danikhani/ITDS-Instructions-WS20 |
As explained in the lecture, a support vector machine is defined by its support vectors. In the sklearn package we can access them and their properties very easily:* support_: indicies of support vectors* support_vectors_: the support vectors* n_support_: the number of support vectors for each class | print('Indicies of support vectors:')
print(classifier.support_)
print('The support vectors:')
print(classifier.support_vectors_)
print('The number of support vectors for each class:')
print(classifier.n_support_) | Indicies of support vectors:
[ 23 24 41 52 56 63 66 68 70 72 76 77 83 84 98 106 110 119
123 126 127 129 133 138 146 147 149]
The support vectors:
[[5.1 3.3 1.7 0.5]
[4.8 3.4 1.9 0.2]
[4.5 2.3 1.3 0.3]
[6.9 3.1 4.9 1.5]
[6.3 3.3 4.7 1.6]
[6.1 2.9 4.7 1.4]
[5.6 3. 4.5 1.5]
[6.2 2.2 4.5 1.5]
[5.9 3... | MIT | Instruction4/Instruction4-RegressionSVM.ipynb | danikhani/ITDS-Instructions-WS20 |
We can also calculate the distance of the data points to the separating hyperplane by using the decision_function(X) method. Score(X,y) calculates the mean accuracy of the classification. The classification report shows metrics such as precision, recall, f1-score and support. You will learn more about these quality met... | from sklearn.metrics import classification_report
classifier.decision_function(descriptiveFeatures_iris)
print('Accuracy: \n', classifier.score(descriptiveFeatures_iris,targetFeature_iris))
print('Classification report: \n')
print(classification_report(targetFeature_iris, targetFeature_iris_predict)) | Accuracy:
0.9933333333333333
Classification report:
precision recall f1-score support
setosa 1.00 1.00 1.00 50
versicolor 1.00 0.98 0.99 50
virginica 0.98 1.00 0.99 50
accuracy 0.99 ... | MIT | Instruction4/Instruction4-RegressionSVM.ipynb | danikhani/ITDS-Instructions-WS20 |
Prepare final dataset | # organize dataset into a useful structure
# create directories
dataset_home = train_folder
# create label subdirectories
labeldirs = ['separate_singleDouble/single/', 'separate_singleDouble/double/']
for labldir in labeldirs:
newdir = dataset_home + labldir
os.makedirs(newdir, exist_ok=True)
# copy train... | _____no_output_____ | MIT | NASA/Python_codes/ML_Books/01_01_transfer_learning_model_EVI.ipynb | HNoorazar/Kirti |
Plot For Fun | # plot dog photos from the dogs vs cats dataset
from matplotlib.image import imread
# define location of dataset
# plot first few images
files = os.listdir(train_folder)[2:4]
# files = [sorted(os.listdir(train_folder))[2]] + [sorted(os.listdir(train_folder))[-2]]
for i in range(2):
# define subplot
pyplot.sub... | _____no_output_____ | MIT | NASA/Python_codes/ML_Books/01_01_transfer_learning_model_EVI.ipynb | HNoorazar/Kirti |
Full Code | # define cnn model
def define_model():
# load model
model = VGG16(include_top=False, input_shape=(224, 224, 3))
# mark loaded layers as not trainable
for layer in model.layers:
layer.trainable = False
# add new classifier layers
flat1 = Flatten()(model.layers[-1].output)
class1 = Den... | _____no_output_____ | MIT | NASA/Python_codes/ML_Books/01_01_transfer_learning_model_EVI.ipynb | HNoorazar/Kirti |
Spark SQLSpark SQL is arguably one of the most important and powerful features in Spark. In a nutshell, with Spark SQL you can run SQL queries against views or tables organized into databases. You also can use system functions or define user functions and analyze query plans in order to optimize their workloads. This ... | spark.sql("SELECT 1 + 1").show() | +-------+
|(1 + 1)|
+-------+
| 2|
+-------+
| MIT | docs/Supplementary-Materials/01-Spark-SQL.ipynb | ymei9/Big-Data-Analytics-for-Business |
As we have seen before, you can completely interoperate between SQL and DataFrames, as you see fit. For instance, you can create a DataFrame, manipulate it with SQL, and then manipulate it again as a DataFrame. It’s a powerful abstraction that you will likely find yourself using quite a bit: | bucket = spark._jsc.hadoopConfiguration().get("fs.gs.system.bucket")
data = "gs://" + bucket + "/notebooks/data/"
spark.read.json(data + "flight-data/json/2015-summary.json")\
.createOrReplaceTempView("flights_view") # DF => SQL
spark.sql("""
SELECT DEST_COUNTRY_NAME, sum(count)
FROM flights_view GROUP BY DEST_COUNT... | _____no_output_____ | MIT | docs/Supplementary-Materials/01-Spark-SQL.ipynb | ymei9/Big-Data-Analytics-for-Business |
Creating TablesYou can create tables from a variety of sources. For instance below we are creating a table from a SELECT statement: | spark.sql('''
CREATE TABLE IF NOT EXISTS flights_from_select USING parquet AS SELECT * FROM flights_view
''')
spark.sql('SELECT * FROM flights_from_select').show(5)
spark.sql('''
DESCRIBE TABLE flights_from_select
''').show() | +-------------------+---------+-------+
| col_name|data_type|comment|
+-------------------+---------+-------+
| DEST_COUNTRY_NAME| string| null|
|ORIGIN_COUNTRY_NAME| string| null|
| count| bigint| null|
+-------------------+---------+-------+
| MIT | docs/Supplementary-Materials/01-Spark-SQL.ipynb | ymei9/Big-Data-Analytics-for-Business |
CatalogThe highest level abstraction in Spark SQL is the Catalog. The Catalog is an abstraction for the storage of metadata about the data stored in your tables as well as other helpful things like databases, tables, functions, and views. The catalog is available in the `spark.catalog` package and contains a number of... | Cat = spark.catalog
Cat.listTables()
spark.sql('SHOW TABLES').show(5, False)
Cat.listDatabases()
spark.sql('SHOW DATABASES').show()
Cat.listColumns('flights_from_select')
Cat.listTables() | _____no_output_____ | MIT | docs/Supplementary-Materials/01-Spark-SQL.ipynb | ymei9/Big-Data-Analytics-for-Business |
Caching Tables | spark.sql('''
CACHE TABLE flights_view
''')
spark.sql('''
UNCACHE TABLE flights_view
''') | _____no_output_____ | MIT | docs/Supplementary-Materials/01-Spark-SQL.ipynb | ymei9/Big-Data-Analytics-for-Business |
Explain | spark.sql('''
EXPLAIN SELECT * FROM just_usa_view
''').show(1, False) | +-----------------------------------------------------------------------------------------------------------------+
|plan |
+---------------------------------------------------------------------------------------... | MIT | docs/Supplementary-Materials/01-Spark-SQL.ipynb | ymei9/Big-Data-Analytics-for-Business |
VIEWS - create/drop | spark.sql('''
CREATE VIEW just_usa_view AS
SELECT * FROM flights_from_select WHERE dest_country_name = 'United States'
''')
spark.sql('''
DROP VIEW IF EXISTS just_usa_view
''') | _____no_output_____ | MIT | docs/Supplementary-Materials/01-Spark-SQL.ipynb | ymei9/Big-Data-Analytics-for-Business |
Drop tables | spark.sql('DROP TABLE flights_from_select')
spark.sql('DROP TABLE IF EXISTS flights_from_select') | _____no_output_____ | MIT | docs/Supplementary-Materials/01-Spark-SQL.ipynb | ymei9/Big-Data-Analytics-for-Business |
Как выложить бота на HEROKU*Подготовил Ян Пиле* Сразу оговоримся, что мы на heroku выкладываем**echo-Бота в телеграме, написанного с помощью библиотеки [pyTelegramBotAPI](https://github.com/eternnoir/pyTelegramBotAPI)**.А взаимодействие его с сервером мы сделаем с использованием [flask](http://flask.pocoo.org/)То есть... | import os
import telebot
from flask import Flask, request
TOKEN = '1403467808:AAEaaLPkIqrhrQ62p7ToJclLtNNINdOopYk' # это мой токен
bot = telebot.TeleBot(token=TOKEN)
server = Flask(__name__)
# Если строка на входе непустая, то бот повторит ее
@bot.message_handler(func=lambda msg: msg.text is not None)
def reply_... | _____no_output_____ | MIT | lect13_NumPy/2021_DPO_13_2_heroku.ipynb | weqrwer/Python_DPO_2021_fall |
Computing Alpha, Beta, and R Squared in Python *Suggested Answers follow (usually there are multiple ways to solve a problem in Python).* *Running a Regression in Python - continued:* | import numpy as np
import pandas as pd
from scipy import stats
import statsmodels.api as sm
import matplotlib.pyplot as plt
data = pd.read_excel('D:/Python/Data_Files/IQ_data.xlsx')
X = data['Test 1']
Y = data['IQ']
plt.scatter(X,Y)
plt.axis([0, 120, 0, 150])
plt.ylabel('IQ')
plt.xlabel('Test 1')
plt.show() | _____no_output_____ | MIT | Python for Finance - Code Files/83 Computing Alpha, Beta, and R Squared in Python/Python 2/Computing Alpha, Beta, and R Squared in Python - Solution.ipynb | siddharthjain1611/Python_for_Finance_Investment_Fundamentals-and-Data-Analytics |
**** Use the statsmodels’ **.add_constant()** method to reassign the X data on X1. Use OLS with arguments Y and X1 and apply the fit method to obtain univariate regression results. Help yourself with the **.summary()** method. | X1 = sm.add_constant(X)
reg = sm.OLS(Y, X1).fit()
reg.summary() | _____no_output_____ | MIT | Python for Finance - Code Files/83 Computing Alpha, Beta, and R Squared in Python/Python 2/Computing Alpha, Beta, and R Squared in Python - Solution.ipynb | siddharthjain1611/Python_for_Finance_Investment_Fundamentals-and-Data-Analytics |
By looking at the p-values, would you conclude Test 1 scores are a good predictor? ***** Imagine a kid would score 84 on Test 1. How many points is she expected to get on the IQ test, approximately? | 45 + 84*0.76 | _____no_output_____ | MIT | Python for Finance - Code Files/83 Computing Alpha, Beta, and R Squared in Python/Python 2/Computing Alpha, Beta, and R Squared in Python - Solution.ipynb | siddharthjain1611/Python_for_Finance_Investment_Fundamentals-and-Data-Analytics |
****** Alpha, Beta, R^2: Apply the stats module’s **linregress()** to extract the value for the slope, the intercept, the r squared, the p_value, and the standard deviation. | slope, intercept, r_value, p_value, std_err = stats.linregress(X,Y)
slope
intercept
r_value
r_value ** 2
p_value
std_err | _____no_output_____ | MIT | Python for Finance - Code Files/83 Computing Alpha, Beta, and R Squared in Python/Python 2/Computing Alpha, Beta, and R Squared in Python - Solution.ipynb | siddharthjain1611/Python_for_Finance_Investment_Fundamentals-and-Data-Analytics |
Use the values of the slope and the intercept to predict the IQ score of a child, who obtained 84 points on Test 1. Is the forecasted value different than the one you obtained above? | intercept + 84 * slope | _____no_output_____ | MIT | Python for Finance - Code Files/83 Computing Alpha, Beta, and R Squared in Python/Python 2/Computing Alpha, Beta, and R Squared in Python - Solution.ipynb | siddharthjain1611/Python_for_Finance_Investment_Fundamentals-and-Data-Analytics |
****** Follow the steps to draw the best fitting line of the provided regression. Define a function that will use the slope and the intercept value to calculate the dots of the best fitting line. | def fitline(b):
return intercept + slope * b | _____no_output_____ | MIT | Python for Finance - Code Files/83 Computing Alpha, Beta, and R Squared in Python/Python 2/Computing Alpha, Beta, and R Squared in Python - Solution.ipynb | siddharthjain1611/Python_for_Finance_Investment_Fundamentals-and-Data-Analytics |
Apply it to the data you have stored in the variable X. | line = fitline(X) | _____no_output_____ | MIT | Python for Finance - Code Files/83 Computing Alpha, Beta, and R Squared in Python/Python 2/Computing Alpha, Beta, and R Squared in Python - Solution.ipynb | siddharthjain1611/Python_for_Finance_Investment_Fundamentals-and-Data-Analytics |
Draw a scatter plot with the X and Y data and then plot X and the obtained fit-line. | plt.scatter(X,Y)
plt.plot(X,line)
plt.show() | _____no_output_____ | MIT | Python for Finance - Code Files/83 Computing Alpha, Beta, and R Squared in Python/Python 2/Computing Alpha, Beta, and R Squared in Python - Solution.ipynb | siddharthjain1611/Python_for_Finance_Investment_Fundamentals-and-Data-Analytics |
# Installs
%%capture
!pip install --upgrade category_encoders plotly
# Imports
import os, sys
os.chdir('/content')
!git init .
!git remote add origin https://github.com/LambdaSchool/DS-Unit-2-Kaggle-Challenge.git
!git pull origin master
!pip install -r requirements.txt
os.chdir('module1')
# Disable warning
import wa... | _____no_output_____ | MIT | Kaggle_Challenge_Assignment_Submission5.ipynb | JimKing100/DS-Unit-2-Kaggle-Challenge | |
Data generators | @numba.njit
def event_series_bernoulli(series_length, event_count):
'''Generate an iid Bernoulli distributed event series.
series_length: length of the event series
event_count: number of events'''
event_series = np.zeros(series_length)
event_series[np.random.choice(np.arange(0, series_length), ev... | _____no_output_____ | MIT | simulations.ipynb | diozaka/eitest |
Visualization of the impact models | default_T = 8192
default_N = 64
default_q = 4
es = event_series_bernoulli(default_T, default_N)
for ts in [
time_series_mean_impact(es, order=default_q, signal_to_noise=10.),
time_series_meanconst_impact(es, order=default_q, const=5.),
time_series_var_impact(es, order=default_q, variance=4.),
time_ser... | _____no_output_____ | MIT | simulations.ipynb | diozaka/eitest |
Simulations | def test_simul_pairs(impact_model, param_T, param_N, param_q, param_r,
n_pairs, lag_cutoff, instantaneous, sample_method,
twosamp_test, multi_test, alpha):
true_positive = 0.
false_positive = 0.
for _ in tqdm(range(n_pairs)):
es = event_series_bernoulli(para... | _____no_output_____ | MIT | simulations.ipynb | diozaka/eitest |
Mean impact model | default_N = 64
default_r = 1.
default_q = 4 | _____no_output_____ | MIT | simulations.ipynb | diozaka/eitest |
... by number of events | vals = [4, 8, 16, 32, 64, 128, 256]
tprs = np.empty(len(vals))
fprs = np.empty(len(vals))
for i, val in enumerate(vals):
tprs[i], fprs[i] = test_simul_pairs(impact_model='mean', param_T=default_T,
param_N=val, param_q=default_q, param_r=default_r,
... | 100%|██████████| 100/100 [00:05<00:00, 18.73it/s]
100%|██████████| 100/100 [00:00<00:00, 451.99it/s]
100%|██████████| 100/100 [00:00<00:00, 439.85it/s]
100%|██████████| 100/100 [00:00<00:00, 379.15it/s]
100%|██████████| 100/100 [00:00<00:00, 276.60it/s]
100%|██████████| 100/100 [00:00<00:00, 163.88it/s]
100%|██████████... | MIT | simulations.ipynb | diozaka/eitest |
... by impact order | vals = [1, 2, 4, 8, 16, 32]
tprs = np.empty(len(vals))
fprs = np.empty(len(vals))
for i, val in enumerate(vals):
tprs[i], fprs[i] = test_simul_pairs(impact_model='mean', param_T=default_T,
param_N=default_N, param_q=val, param_r=default_r,
... | 100%|██████████| 100/100 [00:00<00:00, 218.61it/s]
100%|██████████| 100/100 [00:00<00:00, 187.72it/s]
100%|██████████| 100/100 [00:00<00:00, 207.15it/s]
100%|██████████| 100/100 [00:00<00:00, 200.33it/s]
100%|██████████| 100/100 [00:00<00:00, 213.18it/s]
100%|██████████| 100/100 [00:00<00:00, 215.75it/s]
| MIT | simulations.ipynb | diozaka/eitest |
... by signal-to-noise ratio | vals = [1./32, 1./16, 1./8, 1./4, 1./2, 1., 2., 4.]
tprs = np.empty(len(vals))
fprs = np.empty(len(vals))
for i, val in enumerate(vals):
tprs[i], fprs[i] = test_simul_pairs(impact_model='mean', param_T=default_T,
param_N=default_N, param_q=default_q, param_r=val,
... | 100%|██████████| 100/100 [00:00<00:00, 179.47it/s]
100%|██████████| 100/100 [00:00<00:00, 210.34it/s]
100%|██████████| 100/100 [00:00<00:00, 206.91it/s]
100%|██████████| 100/100 [00:00<00:00, 214.85it/s]
100%|██████████| 100/100 [00:00<00:00, 212.98it/s]
100%|██████████| 100/100 [00:00<00:00, 182.82it/s]
100%|█████████... | MIT | simulations.ipynb | diozaka/eitest |
Meanconst impact model | default_N = 64
default_r = 0.5
default_q = 4 | _____no_output_____ | MIT | simulations.ipynb | diozaka/eitest |
... by number of events | vals = [4, 8, 16, 32, 64, 128, 256]
tprs = np.empty(len(vals))
fprs = np.empty(len(vals))
for i, val in enumerate(vals):
tprs[i], fprs[i] = test_simul_pairs(impact_model='meanconst', param_T=default_T,
param_N=val, param_q=default_q, param_r=default_r,
... | 100%|██████████| 100/100 [00:00<00:00, 370.92it/s]
100%|██████████| 100/100 [00:00<00:00, 387.87it/s]
100%|██████████| 100/100 [00:00<00:00, 364.85it/s]
100%|██████████| 100/100 [00:00<00:00, 313.86it/s]
100%|██████████| 100/100 [00:00<00:00, 215.43it/s]
100%|██████████| 100/100 [00:00<00:00, 115.63it/s]
100%|█████████... | MIT | simulations.ipynb | diozaka/eitest |
... by impact order | vals = [1, 2, 4, 8, 16, 32]
tprs = np.empty(len(vals))
fprs = np.empty(len(vals))
for i, val in enumerate(vals):
tprs[i], fprs[i] = test_simul_pairs(impact_model='meanconst', param_T=default_T,
param_N=default_N, param_q=val, param_r=default_r,
... | 100%|██████████| 100/100 [00:00<00:00, 191.97it/s]
100%|██████████| 100/100 [00:00<00:00, 209.09it/s]
100%|██████████| 100/100 [00:00<00:00, 181.51it/s]
100%|██████████| 100/100 [00:00<00:00, 170.74it/s]
100%|██████████| 100/100 [00:00<00:00, 196.70it/s]
100%|██████████| 100/100 [00:00<00:00, 191.42it/s]
| MIT | simulations.ipynb | diozaka/eitest |
... by mean value | vals = [0.125, 0.25, 0.5, 1, 2]
tprs = np.empty(len(vals))
fprs = np.empty(len(vals))
for i, val in enumerate(vals):
tprs[i], fprs[i] = test_simul_pairs(impact_model='meanconst', param_T=default_T,
param_N=default_N, param_q=default_q, param_r=val,
... | 100%|██████████| 100/100 [00:00<00:00, 172.66it/s]
100%|██████████| 100/100 [00:00<00:00, 212.73it/s]
100%|██████████| 100/100 [00:00<00:00, 210.24it/s]
100%|██████████| 100/100 [00:00<00:00, 153.75it/s]
100%|██████████| 100/100 [00:00<00:00, 211.59it/s]
| MIT | simulations.ipynb | diozaka/eitest |
Variance impact modelIn the paper, we show results with the variance impact model parametrized by the **variance increase**. Here we directly modulate the variance. | default_N = 64
default_r = 8.
default_q = 4 | _____no_output_____ | MIT | simulations.ipynb | diozaka/eitest |
... by number of events | vals = [4, 8, 16, 32, 64, 128, 256]
tprs = np.empty(len(vals))
fprs = np.empty(len(vals))
for i, val in enumerate(vals):
tprs[i], fprs[i] = test_simul_pairs(impact_model='var', param_T=default_T,
param_N=val, param_q=default_q, param_r=default_r,
... | 100%|██████████| 100/100 [00:00<00:00, 379.83it/s]
100%|██████████| 100/100 [00:00<00:00, 399.36it/s]
100%|██████████| 100/100 [00:00<00:00, 372.13it/s]
100%|██████████| 100/100 [00:00<00:00, 319.38it/s]
100%|██████████| 100/100 [00:00<00:00, 216.67it/s]
100%|██████████| 100/100 [00:00<00:00, 121.62it/s]
100%|█████████... | MIT | simulations.ipynb | diozaka/eitest |
... by impact order | vals = [1, 2, 4, 8, 16, 32]
tprs = np.empty(len(vals))
fprs = np.empty(len(vals))
for i, val in enumerate(vals):
tprs[i], fprs[i] = test_simul_pairs(impact_model='var', param_T=default_T,
param_N=default_N, param_q=val, param_r=default_r,
... | 100%|██████████| 100/100 [00:00<00:00, 205.11it/s]
100%|██████████| 100/100 [00:00<00:00, 208.57it/s]
100%|██████████| 100/100 [00:00<00:00, 208.42it/s]
100%|██████████| 100/100 [00:00<00:00, 215.50it/s]
100%|██████████| 100/100 [00:00<00:00, 210.17it/s]
100%|██████████| 100/100 [00:00<00:00, 213.72it/s]
| MIT | simulations.ipynb | diozaka/eitest |
... by variance | vals = [2., 4., 8., 16., 32.]
tprs = np.empty(len(vals))
fprs = np.empty(len(vals))
for i, val in enumerate(vals):
tprs[i], fprs[i] = test_simul_pairs(impact_model='var', param_T=default_T,
param_N=default_N, param_q=default_q, param_r=val,
... | 100%|██████████| 100/100 [00:00<00:00, 211.99it/s]
100%|██████████| 100/100 [00:00<00:00, 213.48it/s]
100%|██████████| 100/100 [00:00<00:00, 209.49it/s]
100%|██████████| 100/100 [00:00<00:00, 214.06it/s]
100%|██████████| 100/100 [00:00<00:00, 213.53it/s]
| MIT | simulations.ipynb | diozaka/eitest |
Tail impact model | default_N = 512
default_r = 3.
default_q = 4 | _____no_output_____ | MIT | simulations.ipynb | diozaka/eitest |
... by number of events | vals = [64, 128, 256, 512, 1024]
tprs = np.empty(len(vals))
fprs = np.empty(len(vals))
for i, val in enumerate(vals):
tprs[i], fprs[i] = test_simul_pairs(impact_model='tail', param_T=default_T,
param_N=val, param_q=default_q, param_r=default_r,
... | 100%|██████████| 100/100 [00:00<00:00, 210.81it/s]
100%|██████████| 100/100 [00:00<00:00, 117.61it/s]
100%|██████████| 100/100 [00:01<00:00, 58.35it/s]
100%|██████████| 100/100 [00:03<00:00, 26.73it/s]
100%|██████████| 100/100 [00:07<00:00, 13.43it/s]
| MIT | simulations.ipynb | diozaka/eitest |
... by impact order | vals = [1, 2, 4, 8, 16, 32]
tprs = np.empty(len(vals))
fprs = np.empty(len(vals))
for i, val in enumerate(vals):
tprs[i], fprs[i] = test_simul_pairs(impact_model='tail', param_T=default_T,
param_N=default_N, param_q=val, param_r=default_r,
... | 100%|██████████| 100/100 [00:03<00:00, 28.23it/s]
100%|██████████| 100/100 [00:03<00:00, 27.89it/s]
100%|██████████| 100/100 [00:03<00:00, 28.22it/s]
100%|██████████| 100/100 [00:03<00:00, 27.32it/s]
100%|██████████| 100/100 [00:03<00:00, 27.25it/s]
100%|██████████| 100/100 [00:03<00:00, 26.63it/s]
| MIT | simulations.ipynb | diozaka/eitest |
... by degrees of freedom | vals = [2.5, 3., 3.5, 4., 4.5, 5., 5.5, 6.]
tprs = np.empty(len(vals))
fprs = np.empty(len(vals))
for i, val in enumerate(vals):
tprs[i], fprs[i] = test_simul_pairs(impact_model='tail', param_T=default_T,
param_N=default_N, param_q=default_q, param_r=val,
... | 100%|██████████| 100/100 [00:03<00:00, 27.68it/s]
100%|██████████| 100/100 [00:03<00:00, 27.97it/s]
100%|██████████| 100/100 [00:03<00:00, 27.91it/s]
100%|██████████| 100/100 [00:03<00:00, 28.07it/s]
100%|██████████| 100/100 [00:03<00:00, 27.99it/s]
100%|██████████| 100/100 [00:03<00:00, 27.71it/s]
100%|██████████| 100... | MIT | simulations.ipynb | diozaka/eitest |
%%capture
%pip install nflfastpy --upgrade
import nflfastpy
from nflfastpy.utils import convert_to_gsis_id
from nflfastpy import default_headshot
from matplotlib import pyplot as plt
import pandas as pd
import seaborn as sns
import requests
print('Example default player headshot\n')
plt.imshow(default_headshot);
df = ... | _____no_output_____ | MIT | examples/Top 25 AY Graph w Roster and Team Logo Data.ipynb | AccidentalGuru/nflfastpy | |
The Analysis of The Evolution of The Russian Comedy. Part 3. In this analysis,we will explore evolution of the French five-act comedy in verse based on the following features:- The coefficient of dialogue vivacity;- The percentage of scenes with split verse lines;- The percentage of scenes with split rhymes;- The perc... | import pandas as pd
import numpy as np
import json
from os import listdir
from scipy.stats import shapiro
import matplotlib.pyplot as plt
%matplotlib inline
import seaborn as sns
def make_plot(feature, title):
mean, std, median = summary(feature)
plt.figure(figsize=(10, 7))
plt.title(title, fontsize=17)
... | _____no_output_____ | MIT | Analyses/The Evolution of The Russian Comedy_Verse_Features.ipynb | innawendell/European_Comedy |
Part 1. Feature Descriptions For the Russian corpus of the five-act comedies, we generated additional features that inspired by Iarkho. So far, we had no understanding how these features evolved over time and whether they could differentiate literary periods. The features include the following:1. **The Coefficient of ... | comedies = pd.read_csv('../Russian_Comedies/Data/Comedies_Raw_Data.csv')
# sort by creation date
comedies_sorted = comedies.sort_values(by='creation_date').copy()
# select only original comedies and five act
original_comedies = comedies_sorted[(comedies_sorted['translation/adaptation'] == 0) &
... | _____no_output_____ | MIT | Analyses/The Evolution of The Russian Comedy_Verse_Features.ipynb | innawendell/European_Comedy |
Part 1. Feature Correlations | comedies_verse_features[['dialogue_vivacity',
'percentage_scene_split_verse',
'percentage_scene_split_rhymes',
'percentage_open_scenes',
'percentage_scenes_rhymes_split_verse']].corr().round(2)
original_comedies[['dialogue_vivac... | _____no_output_____ | MIT | Analyses/The Evolution of The Russian Comedy_Verse_Features.ipynb | innawendell/European_Comedy |
Dialogue vivacity is moderately positively correlated with the percentage of scenes with split verse lines (0.53), with the percentage of scenes with split rhymes (0.51), and slightly less correlated with the percentage of open scenes (0.45). However, it is strongly positively correlated with the percentage of scenes w... | make_plot(comedies_verse_features['dialogue_vivacity'],
'Distribution of the Dialogue Vivacity Coefficient')
mean, std, median = summary(comedies_verse_features['dialogue_vivacity'])
print('Mean dialogue vivacity coefficient', round(mean, 2))
print('Standard deviation of the dialogue vivacity coefficient:', r... | Mean dialogue vivacity coefficient 0.46
Standard deviation of the dialogue vivacity coefficient: 0.1
Median dialogue vivacity coefficient: 0.4575
| MIT | Analyses/The Evolution of The Russian Comedy_Verse_Features.ipynb | innawendell/European_Comedy |
Shapiro-Wilk Normality Test | print('The p-value of the Shapiro-Wilk normality test:',
shapiro(comedies_verse_features['dialogue_vivacity'])[1]) | The p-value of the Shapiro-Wilk normality test: 0.2067030817270279
| MIT | Analyses/The Evolution of The Russian Comedy_Verse_Features.ipynb | innawendell/European_Comedy |
The Shapiro-Wilk test showed that the probability of the coefficient of dialogue vivacity of being normally distributed was 0.2067030817270279, which was above the 0.05 significance level. We failed to reject the null hypothesis of the normal distribution. | make_plot(comedies_verse_features['percentage_scene_split_verse'],
'Distribution of The Percentage of Scenes with Split Verse Lines')
mean, std, median = summary(comedies_verse_features['percentage_scene_split_verse'])
print('Mean percentage of scenes with split verse lines:', round(mean, 2))
print('Standard ... | The p-value of the Shapiro-Wilk normality test: 0.8681985139846802
| MIT | Analyses/The Evolution of The Russian Comedy_Verse_Features.ipynb | innawendell/European_Comedy |
The Shapiro-Wilk showed that the probability of the percentage of scenes with split verse lines of being normally distributed was very high (the p-value is 0.8681985139846802). We failed to reject the null hypothesis of normal distribution. | make_plot(comedies_verse_features['percentage_scene_split_rhymes'],
'Distribution of The Percentage of Scenes with Split Rhymes')
mean, std, median = summary(comedies_verse_features['percentage_scene_split_rhymes'])
print('Mean percentage of scenes with split rhymes:', round(mean, 2))
print('Standard deviation ... | The p-value of the Shapiro-Wilk normality test: 0.5752763152122498
| MIT | Analyses/The Evolution of The Russian Comedy_Verse_Features.ipynb | innawendell/European_Comedy |
The Shapiro-Wilk test showed that the probability of the number of dramatic characters of being normally distributed was 0.5752763152122498. This probability was much higher than the 0.05 significance level. Therefore, we failed to reject the null hypothesis of normal distribution. | make_plot(comedies_verse_features['percentage_open_scenes'],
'Distribution of The Percentage of Open Scenes')
mean, std, median = summary(comedies_verse_features['percentage_open_scenes'])
print('Mean percentage of open scenes:', round(mean, 2))
print('Standard deviation of the percentage of open scenes:', rou... | The p-value of the Shapiro-Wilk normality test: 0.3018988370895386
| MIT | Analyses/The Evolution of The Russian Comedy_Verse_Features.ipynb | innawendell/European_Comedy |
The Shapiro-Wilk test showed that the probability of the number of the percentage of open scenes of being normally distributed was 0.3018988370895386, which was quite a lot higher than the significance level of 0.05. Therefore, we failed to reject the null hypothesis of normal distribution of the percentage of open sce... | make_plot(comedies_verse_features['percentage_scenes_rhymes_split_verse'],
'Distribution of The Percentage of Scenes with Split Verse Lines and Rhymes')
mean, std, median = summary(comedies_verse_features['percentage_scenes_rhymes_split_verse'])
print('Mean percentage of scenes with split rhymes and verse line... | The p-value of the Shapiro-Wilk normality test: 0.015218793414533138
| MIT | Analyses/The Evolution of The Russian Comedy_Verse_Features.ipynb | innawendell/European_Comedy |
The Shapiro-Wilk test showed that the probability of the percentage of scenes with split verse lines and rhymes of being normally distributed was very low (the p-value was 0.015218793414533138). Therefore, we rejected the hypothesis of normal distribution. Summary:1. The majority of the verse features were normally di... | comedies_verse_features['period'] = comedies_verse_features.creation_date.apply(determine_period)
period_one = comedies_verse_features[comedies_verse_features['period'] == 1].copy()
period_two = comedies_verse_features[comedies_verse_features['period'] == 2].copy()
period_one.shape
period_two.shape | _____no_output_____ | MIT | Analyses/The Evolution of The Russian Comedy_Verse_Features.ipynb | innawendell/European_Comedy |
The T-Test The Coefficient of Dialogue Vivacity | from scipy.stats import ttest_ind
ttest_ind(period_one['dialogue_vivacity'],
period_two['dialogue_vivacity'], equal_var=False) | _____no_output_____ | MIT | Analyses/The Evolution of The Russian Comedy_Verse_Features.ipynb | innawendell/European_Comedy |
The Percentage of Scenes With Split Verse Lines | ttest_ind(period_one['percentage_scene_split_verse'],
period_two['percentage_scene_split_verse'], equal_var=False) | _____no_output_____ | MIT | Analyses/The Evolution of The Russian Comedy_Verse_Features.ipynb | innawendell/European_Comedy |
The Percentage of Scnes With Split Rhymes | ttest_ind(period_one['percentage_scene_split_rhymes'],
period_two['percentage_scene_split_rhymes'], equal_var=False) | _____no_output_____ | MIT | Analyses/The Evolution of The Russian Comedy_Verse_Features.ipynb | innawendell/European_Comedy |
The Percentage of Open Scenes | ttest_ind(period_one['percentage_open_scenes'],
period_two['percentage_open_scenes'], equal_var=False) | _____no_output_____ | MIT | Analyses/The Evolution of The Russian Comedy_Verse_Features.ipynb | innawendell/European_Comedy |
Summary|Feature |p-value |Result|---------------------------| ----------------|--------------------------------| The coefficient of dialogue vivacity |0.92 | Not Significant|The percentage of scenes with split verse lines|0.009 | Significant|The percentage of scenes with split rhymes|... | small_sample_mann_whitney_u_test(period_one['percentage_scenes_rhymes_split_verse'],
period_two['percentage_scenes_rhymes_split_verse']) | _____no_output_____ | MIT | Analyses/The Evolution of The Russian Comedy_Verse_Features.ipynb | innawendell/European_Comedy |
Critical Value of U |Periods |Critical Value of U |---------------------------| ----------------| Period One (n=6) and Period Two (n=10) |11 Summary|Feature |u-statistic |Result|---------------------------| ----------------|---------------------------... | def scatter(df, feature, title, xlabel, text_y):
sns.jointplot('creation_date',
feature,
data=df,
color='b',
height=7).plot_joint(
sns.kdeplot,
zorder=0,
n_levels=20)
plt.axvline(1795, color='grey',line... | _____no_output_____ | MIT | Analyses/The Evolution of The Russian Comedy_Verse_Features.ipynb | innawendell/European_Comedy |
The Coefficient of Dialogue Vivacity | scatter(comedies_verse_features,
'dialogue_vivacity',
'The Coefficient of Dialogue Vivacity by Year',
'The Coefficient of Dialogue Vivacity',
0.85) | /opt/anaconda3/envs/text_extraction/lib/python3.7/site-packages/seaborn/_decorators.py:43: FutureWarning: Pass the following variables as keyword args: x, y. From version 0.12, the only valid positional argument will be `data`, and passing other arguments without an explicit keyword will result in an error or misinterp... | MIT | Analyses/The Evolution of The Russian Comedy_Verse_Features.ipynb | innawendell/European_Comedy |
The Percentage of Scenes With Split Verse Lines | scatter(comedies_verse_features,
'percentage_scene_split_verse',
'The Percentage of Scenes With Split Verse Lines by Year',
'Percentage of Scenes With Split Verse Lines',
80) | /opt/anaconda3/envs/text_extraction/lib/python3.7/site-packages/seaborn/_decorators.py:43: FutureWarning: Pass the following variables as keyword args: x, y. From version 0.12, the only valid positional argument will be `data`, and passing other arguments without an explicit keyword will result in an error or misinterp... | MIT | Analyses/The Evolution of The Russian Comedy_Verse_Features.ipynb | innawendell/European_Comedy |
The Percentage of Scenes With Split Rhymes | scatter(comedies_verse_features,
'percentage_scene_split_rhymes',
'The Percentage of Scenes With Split Rhymes by Year',
'The Percentage of Scenes With Split Rhymes',
80) | /opt/anaconda3/envs/text_extraction/lib/python3.7/site-packages/seaborn/_decorators.py:43: FutureWarning: Pass the following variables as keyword args: x, y. From version 0.12, the only valid positional argument will be `data`, and passing other arguments without an explicit keyword will result in an error or misinterp... | MIT | Analyses/The Evolution of The Russian Comedy_Verse_Features.ipynb | innawendell/European_Comedy |
The Percentage of Open Scenes | scatter(comedies_verse_features,
'percentage_open_scenes',
'The Percentage of Open Scenes by Year',
'The Percentage of Open Scenes',
100) | /opt/anaconda3/envs/text_extraction/lib/python3.7/site-packages/seaborn/_decorators.py:43: FutureWarning: Pass the following variables as keyword args: x, y. From version 0.12, the only valid positional argument will be `data`, and passing other arguments without an explicit keyword will result in an error or misinterp... | MIT | Analyses/The Evolution of The Russian Comedy_Verse_Features.ipynb | innawendell/European_Comedy |
The Percentage of Scenes With Split Verse Lines and Rhymes | scatter(comedies_verse_features,
'percentage_scenes_rhymes_split_verse',
' The Percentage of Scenes With Split Verse Lines and Rhymes by Year',
' The Percentage of Scenes With Split Verse Lines and Rhymes',
45) | /opt/anaconda3/envs/text_extraction/lib/python3.7/site-packages/seaborn/_decorators.py:43: FutureWarning: Pass the following variables as keyword args: x, y. From version 0.12, the only valid positional argument will be `data`, and passing other arguments without an explicit keyword will result in an error or misinterp... | MIT | Analyses/The Evolution of The Russian Comedy_Verse_Features.ipynb | innawendell/European_Comedy |
Part 5. Descriptive Statistics For Two Periods and Overall The Coefficient of Dialogue Vivacity In Entire Corpus | comedies_verse_features.describe().loc[:, 'dialogue_vivacity'][['mean',
'std',
'50%',
'min',
... | _____no_output_____ | MIT | Analyses/The Evolution of The Russian Comedy_Verse_Features.ipynb | innawendell/European_Comedy |
By Tentative Periods | comedies_verse_features.groupby('period').describe().loc[:, 'dialogue_vivacity'][['mean',
'std',
'50%',
... | _____no_output_____ | MIT | Analyses/The Evolution of The Russian Comedy_Verse_Features.ipynb | innawendell/European_Comedy |
The Percentage of Scenes With Split Verse Lines In Entire Corpus | comedies_verse_features.describe().loc[:, 'percentage_scene_split_verse'][['mean',
'std',
'50%',
'min',
... | _____no_output_____ | MIT | Analyses/The Evolution of The Russian Comedy_Verse_Features.ipynb | innawendell/European_Comedy |
By Periods | comedies_verse_features.groupby('period').describe().loc[:, 'percentage_scene_split_verse'][['mean',
'std',
'50%',
... | _____no_output_____ | MIT | Analyses/The Evolution of The Russian Comedy_Verse_Features.ipynb | innawendell/European_Comedy |
The Percentage of Scenes With Split Rhymes | comedies_verse_features.describe().loc[:, 'percentage_scene_split_rhymes'][['mean',
'std',
'50%',
... | _____no_output_____ | MIT | Analyses/The Evolution of The Russian Comedy_Verse_Features.ipynb | innawendell/European_Comedy |
By Tentative Periods | comedies_verse_features.groupby('period').describe().loc[:, 'percentage_scene_split_rhymes'][['mean',
'std',
'50%',
... | _____no_output_____ | MIT | Analyses/The Evolution of The Russian Comedy_Verse_Features.ipynb | innawendell/European_Comedy |
The Percentage of Open Scenes In Entire Corpus | comedies_verse_features.describe().loc[:, 'percentage_open_scenes'][['mean',
'std',
'50%',
'min',
... | _____no_output_____ | MIT | Analyses/The Evolution of The Russian Comedy_Verse_Features.ipynb | innawendell/European_Comedy |
By Tenative Periods | comedies_verse_features.groupby('period').describe().loc[:, 'percentage_open_scenes'][['mean',
'std',
'50%',
... | _____no_output_____ | MIT | Analyses/The Evolution of The Russian Comedy_Verse_Features.ipynb | innawendell/European_Comedy |
The Percentage of Scenes With Split Verse Lines or Rhymes | comedies_verse_features.describe().loc[:, 'percentage_scenes_rhymes_split_verse'][['mean',
'std',
'50%',
... | _____no_output_____ | MIT | Analyses/The Evolution of The Russian Comedy_Verse_Features.ipynb | innawendell/European_Comedy |
Lalonde Pandas API Exampleby Adam Kelleher We'll run through a quick example using the high-level Python API for the DoSampler. The DoSampler is different from most classic causal effect estimators. Instead of estimating statistics under interventions, it aims to provide the generality of Pearlian causal inference. In... | import os, sys
sys.path.append(os.path.abspath("../../../"))
from rpy2.robjects import r as R
%load_ext rpy2.ipython
#%R install.packages("Matching")
%R library(Matching)
%R data(lalonde)
%R -o lalonde
lalonde.to_csv("lalonde.csv",index=False)
# the data already loaded in the previous cell. we include the import
# her... | _____no_output_____ | MIT | Utils/dowhy/docs/source/example_notebooks/lalonde_pandas_api.ipynb | maliha93/Fairness-Analysis-Code |
The `causal` Namespace We've created a "namespace" for `pandas.DataFrame`s containing causal inference methods. You can access it here with `lalonde.causal`, where `lalonde` is our `pandas.DataFrame`, and `causal` contains all our new methods! These methods are magically loaded into your existing (and future) datafram... | import dowhy.api | _____no_output_____ | MIT | Utils/dowhy/docs/source/example_notebooks/lalonde_pandas_api.ipynb | maliha93/Fairness-Analysis-Code |
Now that we have the `causal` namespace, lets give it a try! The `do` OperationThe key feature here is the `do` method, which produces a new dataframe replacing the treatment variable with values specified, and the outcome with a sample from the interventional distribution of the outcome. If you don't specify a value ... | do_df = lalonde.causal.do(x='treat',
outcome='re78',
common_causes=['nodegr', 'black', 'hisp', 'age', 'educ', 'married'],
variable_types={'age': 'c', 'educ':'c', 'black': 'd', 'hisp': 'd',
'married':... | WARNING:dowhy.causal_model:Causal Graph not provided. DoWhy will construct a graph based on data inputs.
INFO:dowhy.causal_graph:If this is observed data (not from a randomized experiment), there might always be missing confounders. Adding a node named "Unobserved Confounders" to reflect this.
INFO:dowhy.causal_model:M... | MIT | Utils/dowhy/docs/source/example_notebooks/lalonde_pandas_api.ipynb | maliha93/Fairness-Analysis-Code |
Notice you get the usual output and prompts about identifiability. This is all `dowhy` under the hood!We now have an interventional sample in `do_df`. It looks very similar to the original dataframe. Compare them: | lalonde.head()
do_df.head() | _____no_output_____ | MIT | Utils/dowhy/docs/source/example_notebooks/lalonde_pandas_api.ipynb | maliha93/Fairness-Analysis-Code |
Treatment Effect EstimationWe could get a naive estimate before for a treatment effect by doing | (lalonde[lalonde['treat'] == 1].mean() - lalonde[lalonde['treat'] == 0].mean())['re78'] | _____no_output_____ | MIT | Utils/dowhy/docs/source/example_notebooks/lalonde_pandas_api.ipynb | maliha93/Fairness-Analysis-Code |
We can do the same with our new sample from the interventional distribution to get a causal effect estimate | (do_df[do_df['treat'] == 1].mean() - do_df[do_df['treat'] == 0].mean())['re78'] | _____no_output_____ | MIT | Utils/dowhy/docs/source/example_notebooks/lalonde_pandas_api.ipynb | maliha93/Fairness-Analysis-Code |
We could get some rough error bars on the outcome using the normal approximation for a 95% confidence interval, like | import numpy as np
1.96*np.sqrt((do_df[do_df['treat'] == 1].var()/len(do_df[do_df['treat'] == 1])) +
(do_df[do_df['treat'] == 0].var()/len(do_df[do_df['treat'] == 0])))['re78'] | _____no_output_____ | MIT | Utils/dowhy/docs/source/example_notebooks/lalonde_pandas_api.ipynb | maliha93/Fairness-Analysis-Code |
but note that these DO NOT contain propensity score estimation error. For that, a bootstrapping procedure might be more appropriate. This is just one statistic we can compute from the interventional distribution of `'re78'`. We can get all of the interventional moments as well, including functions of `'re78'`. We can l... | do_df['re78'].describe()
lalonde['re78'].describe() | _____no_output_____ | MIT | Utils/dowhy/docs/source/example_notebooks/lalonde_pandas_api.ipynb | maliha93/Fairness-Analysis-Code |
and even plot aggregations, like | %matplotlib inline
import seaborn as sns
sns.barplot(data=lalonde, x='treat', y='re78')
sns.barplot(data=do_df, x='treat', y='re78') | _____no_output_____ | MIT | Utils/dowhy/docs/source/example_notebooks/lalonde_pandas_api.ipynb | maliha93/Fairness-Analysis-Code |
Specifying InterventionsYou can find the distribution of the outcome under an intervention to set the value of the treatment. | do_df = lalonde.causal.do(x={'treat': 1},
outcome='re78',
common_causes=['nodegr', 'black', 'hisp', 'age', 'educ', 'married'],
variable_types={'age': 'c', 'educ':'c', 'black': 'd', 'hisp': 'd',
'marr... | _____no_output_____ | MIT | Utils/dowhy/docs/source/example_notebooks/lalonde_pandas_api.ipynb | maliha93/Fairness-Analysis-Code |
This new dataframe gives the distribution of `'re78'` when `'treat'` is set to `1`. For much more detail on how the `do` method works, check the docstring: | help(lalonde.causal.do) | Help on method do in module dowhy.api.causal_data_frame:
do(x, method='weighting', num_cores=1, variable_types={}, outcome=None, params=None, dot_graph=None, common_causes=None, estimand_type='nonparametric-ate', proceed_when_unidentifiable=False, stateful=False) method of dowhy.api.causal_data_frame.CausalAccessor in... | MIT | Utils/dowhy/docs/source/example_notebooks/lalonde_pandas_api.ipynb | maliha93/Fairness-Analysis-Code |
Welcome to the Datenguide Python PackageWithin this notebook the functionality of the package will be explained and demonstrated with examples. Topics- Import- get region IDs- get statstic IDs- get the data - for single regions - for multiple regions 1. Import **Import the helper functions 'get_all_regions' and... | # ONLY FOR TESTING LOCAL PACKAGE
# %cd ..
from datenguidepy.query_helper import get_all_regions, get_statistics
from datenguidepy import Query | C:\Users\Alexandra\Documents\GitHub\datenguide-python
| MIT | use_case/01_intro_tutorial.ipynb | elekt/datenguide-python |
**Import pandas and matplotlib for the usual display of data as tables and graphs** | import pandas as pd
import matplotlib
%matplotlib inline
pd.set_option('display.max_colwidth', 150) | _____no_output_____ | MIT | use_case/01_intro_tutorial.ipynb | elekt/datenguide-python |
2. Get Region IDs How to get the ID of the region I want to query Regionalstatistik - the database behind Datenguide - has data for differently granular levels of Germany. nuts: 1 – Bundesländer 2 – Regierungsbezirke / statistische Regionen 3 – Kreise / kreisfreie Städte. lau: 1 -... | # get_all_regions returns all ids
get_all_regions() | _____no_output_____ | MIT | use_case/01_intro_tutorial.ipynb | elekt/datenguide-python |
To get a specific ID, use the common pandas function `query()` | # e.g. get all "Bundesländer
get_all_regions().query("level == 'nuts1'")
# e.g. get the ID of Havelland
get_all_regions().query("name =='Havelland'") | _____no_output_____ | MIT | use_case/01_intro_tutorial.ipynb | elekt/datenguide-python |
3. Get statistic IDs How to find statistics | # get all statistics
get_statistics() | _____no_output_____ | MIT | use_case/01_intro_tutorial.ipynb | elekt/datenguide-python |
If you already know the statsitic ID you are looking for - perfect. Otherwise you can use the pandas `query()` function so search e.g. for specific terms. | # find out the name of the desired statistic about birth
get_statistics().query('long_description.str.contains("Statistik der Geburten")', engine='python') | _____no_output_____ | MIT | use_case/01_intro_tutorial.ipynb | elekt/datenguide-python |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.