Open Class Python¶
Python ITB
Makers Institute
13 February 2018
About Us¶
Since 12 January 2016
First open class: September 2016
About Me¶
Aris Budi Wibowo¶
Outline¶
You can also view this repository from: https://github.com/arisbw/open-class-makers
Introduction to Python¶
Say hello to dat snake. 🐍
In [1]:
print("Hello Python!")
Object types in Python¶
Number¶
If you forgot what the simple math looks like...
In [2]:
1 + 1
Out[2]:
In [3]:
8 - 1
Out[3]:
In [4]:
10 * 2
Out[4]:
In [5]:
35 / 5
Out[5]:
In [6]:
5 // 3
Out[6]:
In [7]:
5.0 // 3.0
Out[7]:
In [2]:
2**4
Out[2]:
String¶
In [28]:
s = "ayam"
len(s)
Out[28]:
In [29]:
s[0]
Out[29]:
Or, we can do this!
In [30]:
s[-1]
Out[30]:
In [31]:
s[-2]
Out[31]:
We can also slice that string
In [32]:
s[0:3]
Out[32]:
In [33]:
s[1:]
Out[33]:
In [34]:
s[:5]
Out[34]:
In [35]:
s[:-1]
Out[35]:
Or add with another string
In [36]:
s + ' ' +'euy'
Out[36]:
In [37]:
s*4
Out[37]:
Find and replace substring
In [39]:
s.find('yam')
Out[39]:
In [40]:
s.replace('ayam', 'bebek')
Out[40]:
In [42]:
line = 'aaa,bbb,ccccc,dd'
In [43]:
line.split(',')
Out[43]:
In [41]:
s
Out[41]:
In [44]:
s.upper()
Out[44]:
In [50]:
s.isalpha()
Out[50]:
List¶
In [1]:
L = [123, 'spam', 1.23]
In [2]:
L[0]
Out[2]:
In [3]:
L[:-1]
Out[3]:
In [4]:
L + [4, 5, 6]
Out[4]:
In [5]:
L
Out[5]:
In [6]:
L.append('hehe')
In [7]:
L
Out[7]:
In [8]:
L.pop(1)
Out[8]:
In [9]:
L
Out[9]:
Dictionaries¶
In [117]:
D = {'food': 'Spam', 'quantity': 4, 'color': 'pink'}
In [118]:
D['food']
Out[118]:
In [119]:
D['quantity']
Out[119]:
In [120]:
D = {}
D['name'] = 'Bob'# Create keys by assignment
D['job'] = 'dev'
D['age'] = 40
In [121]:
D
Out[121]:
What if...¶
In [8]:
a = 90
if a > 50:
print("Lebih dari 50")
else:
print("Kurang dari 50")
In [9]:
a = 3
"seeep" if a > 2 else 2
Out[9]:
Now, what ifs...
In [10]:
a = 400
if a < 100:
print("Kurang dari 100")
elif a >= 100 and a <= 500:
print("Antara 100 dan 500")
else:
print("Lebih dari 500")
... say that again?
In [11]:
a = 5
for i in range(a):
if a > 10:
print("Lebih dari 10")
else:
print("Kurang dari 10")
or
In [12]:
a = 0
while a < 5:
print("Gampang euy")
a += 1
Functions!
In [13]:
def tambah(a=0,b=0):
c = a+b
return c
tambah(2,3)
Out[13]:
Pandas!¶
Initialize libraries that we will use later.
In [1]:
import pandas as pd
import matplotlib.pyplot as plt
%matplotlib inline
First, let's download the data we need.
In [2]:
!curl https://raw.githubusercontent.com/fivethirtyeight/data/master/bob-ross/elements-by-episode.csv -O elements-by-episode.csv
Or you can use wget if you use linux.
In [15]:
df = pd.read_csv("elements-by-episode.csv")
Let's peek the data.
In [16]:
df.head()
Out[16]:
In [17]:
df.columns
Out[17]:
In [18]:
df.shape
Out[18]:
Summary of the data:
In [19]:
df.describe()
Out[19]:
Transpose the data
In [20]:
df.T.head()
Out[20]:
Sort the data
In [26]:
df.sort_values(by='TREES',ascending=False).head()
Out[26]:
Select and Slice
In [21]:
df.EPISODE[1:5]
Out[21]:
In [22]:
df['EPISODE'][1:5]
Out[22]:
In [23]:
df.loc[0:2]
Out[23]:
In [29]:
df.loc[100:106,['WATERFALL','TREE']]
Out[29]:
In [37]:
df[(df.BRIDGE>0) & (df.WINTER>0)]
Out[37]:
In [82]:
df1 = df.iloc[:,3:6]
In [83]:
df1.head()
Out[83]:
In [84]:
df1.plot.hist()
Out[84]:
In [95]:
df1.iloc[30:100,:].plot(figsize=(8, 8))
Out[95]:
a bit Intro to Machine Learning¶
What is Machine Learning?¶
Machine Learning: ... a method of teaching computers to make and improve predictions or behaviors based on some data.
Basically, there are 3* main task in machine learning:
- Supervised learning: your examples must be labeled
- Unsupervised learning: your examples are not labeled
- Reinforcement learning: based on reward and punishment
Why Machine Learning?¶
I mean... why now?
So many algorithms that are really good right now:
- SVM
- Tree-based algorithms: Random forest, XGboost, catboost
- Deep learning
- etc.
a bit of KNN¶
DISCLAIMER: This is the oversimplified version
A classification and regression method. In this case, we will explore in the classification problem.
In [97]:
from sklearn import neighbors
import numpy as np
%matplotlib inline
import seaborn
Create dataset:
In [98]:
training_data = pd.DataFrame()
training_data['test_1'] = [0.3051,0.4949,0.6974,0.3769,0.2231,0.341,0.4436,0.5897,0.6308,0.5]
training_data['test_2'] = [0.5846,0.2654,0.2615,0.4538,0.4615,0.8308,0.4962,0.3269,0.5346,0.6731]
training_data['outcome'] = ['win','win','win','win','win','loss','loss','loss','loss','loss']
training_data.head()
Out[98]:
Plot the data
In [99]:
seaborn.lmplot('test_1', 'test_2', data=training_data, fit_reg=False,hue="outcome", scatter_kws={"marker": "D","s": 100})
Out[99]:
Convert data to np.arrays
In [100]:
X = training_data.as_matrix(columns=['test_1', 'test_2'])
y = np.array(training_data['outcome'])
Training!
In [101]:
clf = neighbors.KNeighborsClassifier(3, weights = 'uniform')
trained_model = clf.fit(X, y)
drum rolls 🥁
In [102]:
trained_model.score(X, y)
Out[102]:
Check the model result to new data point
In [103]:
x_test = np.array([[.4,.6]])
In [104]:
trained_model.predict(x_test)
Out[104]:
Check result
In [105]:
trained_model.predict_proba(x_test)
Out[105]:
Best k?¶
Load libraries
In [109]:
from sklearn.neighbors import KNeighborsClassifier
from sklearn import datasets
from sklearn.preprocessing import StandardScaler
from sklearn.pipeline import Pipeline, FeatureUnion
from sklearn.model_selection import GridSearchCV
Load Iris Flower Data
In [110]:
iris = datasets.load_iris()
X = iris.data
y = iris.target
Standardize Data
In [111]:
standardizer = StandardScaler()
X_std = standardizer.fit_transform(X)
Create knn model
In [112]:
knn = KNeighborsClassifier(n_neighbors=5, metric='euclidean', n_jobs=-1).fit(X_std, y)
Create search space of possible value of k
In [113]:
pipe = Pipeline([('standardizer', standardizer), ('knn', knn)])
search_space = [{'knn__n_neighbors': [1, 2, 3, 4, 5, 6, 7, 8, 9, 10]}]
Search!
In [115]:
clf = GridSearchCV(pipe, search_space, cv=5, verbose=0).fit(X_std, y)
Show best k
In [116]:
clf.best_estimator_.get_params()['knn__n_neighbors']
Out[116]:
References¶
Thank you! 🙇🙏¶
Misc.¶
In [1]:
%mkdir image\paintings
%cd image\paintings
In [2]:
#Credit to: Jeff Thompson (https://gist.github.com/jeffThompson/6d4c45f89dc907925775972e72d9cf00)
num_images=403
import urllib.request
from tqdm import *
for i in tqdm(range(1, num_images+1)):
try:
url = 'http://www.twoinchbrush.com/images/painting' + str(i) + '.png'
filename = ('%03d' % (i,)) + '.png'
urllib.request.urlretrieve(url, filename)
except:
print('- ERROR!')
In [ ]: