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 |
|---|---|---|---|---|---|
Gym Crowdedness Analysis with PCA > Objective : To **predict** how crowded a university gym would be at a given time of day (and some other features, including weather) > Data Decription : The dataset consists of 26,000 people counts (about every 10 minutes) over one year. The dataset also contains information abo... | import numpy as np # linear algebra
import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns
%matplotlib inline
df=pd.read_csv(r'C:\Users\kusht\OneDrive\Desktop\Excel-csv\PCA analysis.csv') #Replace it with your path where the data file is stored
df.head() | _____no_output_____ | MIT | Gym Crowd Analysis/Gym Crowd Analysis with PCA (ToDo Template).ipynb | abhisngh/Data-Science |
**TASK : Print the `info()` of the dataset** | ### START CODE HERE (~ 1 Line of code)
### END CODE | _____no_output_____ | MIT | Gym Crowd Analysis/Gym Crowd Analysis with PCA (ToDo Template).ipynb | abhisngh/Data-Science |
**TASK : Describe the dataset using `describe()`** | ### START CODE HERE (~ 1 Line of code)
### END CODE | _____no_output_____ | MIT | Gym Crowd Analysis/Gym Crowd Analysis with PCA (ToDo Template).ipynb | abhisngh/Data-Science |
**TASK : Convert temperature in farenheit into celsius scale using the formula `Celsius=(Fahrenheit-32)* (5/9)`** | ### START CODE HERE (~1 Line of code)
### END CODE | _____no_output_____ | MIT | Gym Crowd Analysis/Gym Crowd Analysis with PCA (ToDo Template).ipynb | abhisngh/Data-Science |
**TASK : Convert the timestamp into hours in 12 h format as its currently in seconds and drop `date` coulmn** | ### START CODE HERE: (~ 1 Line of code)
### END CODE | _____no_output_____ | MIT | Gym Crowd Analysis/Gym Crowd Analysis with PCA (ToDo Template).ipynb | abhisngh/Data-Science |
`2` Exploratory Data Analysis `2.1` Uni-Variate and Bi-Variate Analysis - **Pair Plots** **TASK : Use `pairplot()` to make different pair scatter plots of the entire dataframe** | ### START CODE HERE :
### END CODE | _____no_output_____ | MIT | Gym Crowd Analysis/Gym Crowd Analysis with PCA (ToDo Template).ipynb | abhisngh/Data-Science |
**TASK: Now analyse scatter plots between `number_people` and all other attributes using a `for loop` to properly know what are the ideal conditions for people to come to the gym** | ### START CODE HERE
### END CODE | _____no_output_____ | MIT | Gym Crowd Analysis/Gym Crowd Analysis with PCA (ToDo Template).ipynb | abhisngh/Data-Science |
**Analyse the plots and understand :**1. **At what time , temperature , week of the day more people come in?** 2. **Whether people like to come to the gym in a holiday or a weekend or they prefer to come to gym during working days?** 3. **Which month is most preferable for people to come to the gym?** - *... | ### START CODE HERE :
### END CODE | _____no_output_____ | MIT | Gym Crowd Analysis/Gym Crowd Analysis with PCA (ToDo Template).ipynb | abhisngh/Data-Science |
`2.2` Correlation Matrix **TASK : Plot a correlation matrix and make it more understandable using `sns.heatmap`** | ### START CODE HERE :
### END CODE HERE | _____no_output_____ | MIT | Gym Crowd Analysis/Gym Crowd Analysis with PCA (ToDo Template).ipynb | abhisngh/Data-Science |
**Analyse the correlation matrix and understand the different dependencies of attributes on each other** `3.` Processing : `3.1` One hot encoding :One hot encoding certain attributes to not give any ranking/priority to any instance **TASK: One Hot Encode following attributes `month` , `hour` , `day of week`** | ## YOU CAN USE EITHER get_dummies() OR OneHotEncoder()
### START CODE HERE :
### END CODE | _____no_output_____ | MIT | Gym Crowd Analysis/Gym Crowd Analysis with PCA (ToDo Template).ipynb | abhisngh/Data-Science |
`3.2` Feature Scaling :Some attributes ranges are ver different compared to other values and during PCA implementation this might give a problem thus you need to standardise some of the attributes **TASK: Using `StandardScaler()` , standardise `temperature` and `timestamp`** | ## You can use two individual scalers one for temperature and other for timestamp
## you can use an array type data=df.values and standradise data then split data into X and y
from sklearn.preprocessing import StandardScaler
### START CODE HERE : (Replace places having '#' with the code)
data=df.values
scaler1 = Standa... | _____no_output_____ | MIT | Gym Crowd Analysis/Gym Crowd Analysis with PCA (ToDo Template).ipynb | abhisngh/Data-Science |
`4.` Splitting the dataset : **TASK : Split the dataset into dependent and independent variables and name them y and X respectively** | ### START CODE HERE :
### END CODE | _____no_output_____ | MIT | Gym Crowd Analysis/Gym Crowd Analysis with PCA (ToDo Template).ipynb | abhisngh/Data-Science |
**TASK : Split the X ,y into training and test set** | from sklearn.model_selection import train_test_split
### START CODE HERE :
### END CODE | _____no_output_____ | MIT | Gym Crowd Analysis/Gym Crowd Analysis with PCA (ToDo Template).ipynb | abhisngh/Data-Science |
`5.` Principal Component Analysis Principal component analysis (PCA) is a technique for reducing the dimensionality of such datasets, increasing interpretability but at the same time minimizing information loss. It does so by creating new uncorrelated variables that successively maximize variance. **How does it work? ... | from sklearn.decomposition import PCA
### START CODE HERE : (Replace spaces having '#' with the code)
pca = PCA()
pca.fit_transform(#)
### END CODE | _____no_output_____ | MIT | Gym Crowd Analysis/Gym Crowd Analysis with PCA (ToDo Template).ipynb | abhisngh/Data-Science |
**TASK : Get covariance using `get_covariance()`** | ### START CODE HERE (~ 1 line of code)
### END CODE HERE | _____no_output_____ | MIT | Gym Crowd Analysis/Gym Crowd Analysis with PCA (ToDo Template).ipynb | abhisngh/Data-Science |
**TASK : Get explained variance using `explained_variance_ratio`** | ### START CODE HERE :
### END CODE | _____no_output_____ | MIT | Gym Crowd Analysis/Gym Crowd Analysis with PCA (ToDo Template).ipynb | abhisngh/Data-Science |
**TASK : Plot a bar graph of `explained variance`** | # you can use plt.bar()
### START CODE HERE : (Replace spaces having '#' with the code)
with plt.style.context('dark_background'):
plt.figure(figsize=(15,12))
plt.bar(range(49), '#', alpha=0.5, align='center',
label='individual explained variance')
plt.ylabel('#')
plt.xlabel('#')
plt.l... | _____no_output_____ | MIT | Gym Crowd Analysis/Gym Crowd Analysis with PCA (ToDo Template).ipynb | abhisngh/Data-Science |
**Analyse the plot and estimate how many componenets you want to keep** **TASK : Make a `PCA()` object with n_components =20 and fit-transform in the dataset (X) and assign to a new variable `X_new`** | ### START CODE HERE :
### END CODE | _____no_output_____ | MIT | Gym Crowd Analysis/Gym Crowd Analysis with PCA (ToDo Template).ipynb | abhisngh/Data-Science |
Now , `X_new` is the dataset for PCA **TASK : Get Covariance using `get_covariance`** | ### START CODE HERE (~1 Line of code)
### END CODE | _____no_output_____ | MIT | Gym Crowd Analysis/Gym Crowd Analysis with PCA (ToDo Template).ipynb | abhisngh/Data-Science |
**TASK : Get the explained variance using `explained_variance_ratio`** | ### START CODE HERE :
### END CODE | _____no_output_____ | MIT | Gym Crowd Analysis/Gym Crowd Analysis with PCA (ToDo Template).ipynb | abhisngh/Data-Science |
**TASK : Plot bar plot of `exlpained variance`** | # You can use plt.bar()
### START CODE HERE:
### END CODE | _____no_output_____ | MIT | Gym Crowd Analysis/Gym Crowd Analysis with PCA (ToDo Template).ipynb | abhisngh/Data-Science |
`6.` Modelling : Random Forest To understand Random forest classifier , lets first get a brief idea about Decision Trees in general. Decision Trees are very intuitive and at everyone have used this knowingly or unknowingly at some point . Basically the model keeps sorting them into categories forming a large tree by r... | # Establish model
from sklearn.ensemble import RandomForestRegressor
model = RandomForestRegressor()
# Try different numbers of n_estimators and print the scores
# You can use a variable estimators = np.arrange(10,200,10) and then a for loop to take all the values of estimators
### START CODE HERE : (Replace spaces ha... | _____no_output_____ | MIT | Gym Crowd Analysis/Gym Crowd Analysis with PCA (ToDo Template).ipynb | abhisngh/Data-Science |
**TASK : Make a plot between `n_estimator` and `scores` to properly get the best number of estimators** | ## Use plt.plot
### START CODE HERE :
### END CODE HERE | _____no_output_____ | MIT | Gym Crowd Analysis/Gym Crowd Analysis with PCA (ToDo Template).ipynb | abhisngh/Data-Science |
`6.2` Random Forest With PCA **TASK : Split the your dataset with PCA into training and testing set using `train_test_split`** | from sklearn.model_selection import train_test_split
### START CODE HERE :
### END CODE | _____no_output_____ | MIT | Gym Crowd Analysis/Gym Crowd Analysis with PCA (ToDo Template).ipynb | abhisngh/Data-Science |
**TASK : Make a random forest model called `model_pca` and fit it into the new X_train and y_train and then print out the random forest scores for dataset with PCA applied to it** | # Establish model
from sklearn.ensemble import RandomForestRegressor
model_pca = RandomForestRegressor()
# You can use different number of estimators
# # You can use a variable estimators = np.arrange(10,200,10) and then a for loop to take all the values of estimators
### START CODE HERE :
### END CODE | _____no_output_____ | MIT | Gym Crowd Analysis/Gym Crowd Analysis with PCA (ToDo Template).ipynb | abhisngh/Data-Science |
**TASK : Make a plot between `n_estimator` and `score` and find the best parameter** | # you can use plt.plot
### START CODE HERE :
### END CODE | _____no_output_____ | MIT | Gym Crowd Analysis/Gym Crowd Analysis with PCA (ToDo Template).ipynb | abhisngh/Data-Science |
Performance programming We've spent most of this course looking at how to make code readable and reliable. For research work, it is often also important that code is efficient: that it does what it needs to do *quickly*. It is very hard to work out beforehand whether code will be efficient or not: it is essential to *... | def mandel1(position, limit=50):
value = position
while abs(value) < 2:
limit -= 1
value = value**2 + position
if limit < 0:
return 0
return limit
xmin = -1.5
ymin = -1.0
xmax = 0.5
ymax = 1.0
resolution = 300
xstep = (xmax - xmin) / resolution
... | _____no_output_____ | CC-BY-3.0 | ch08performance/010intro.ipynb | jack89roberts/rsd-engineeringcourse |
We will learn this lesson how to make a version of this code which works Ten Times faster: | import numpy as np
def mandel_numpy(position,limit=50):
value = position
diverged_at_count = np.zeros(position.shape)
while limit > 0:
limit -= 1
value = value**2+position
diverging = value * np.conj(value) > 4
first_diverged_this_time = np.logical_and(diverging, diverged_at_... | 50.9 ms ± 10.8 ms per loop (mean ± std. dev. of 7 runs, 10 loops each)
| CC-BY-3.0 | ch08performance/010intro.ipynb | jack89roberts/rsd-engineeringcourse |
Note we get the same answer: | sum(sum(abs(data_numpy - data1))) | _____no_output_____ | CC-BY-3.0 | ch08performance/010intro.ipynb | jack89roberts/rsd-engineeringcourse |
Matplotlib ( Matplotlib Pt. 3) Plot Appearence in Matplotlib | import matplotlib.pyplot as plt
%matplotlib inline
import numpy as np
x = np.linspace(0,5,11) # We go from 0 to 5 and grab 11 points which are linearly spaced.
y = x ** 2
fig = plt.figure()
# Add a set of axes to the figure.
ax = fig.add_axes([0,0,1,1])
# To add color to the plot there are multiple ways like directly t... | _____no_output_____ | BSD-3-Clause | 08. Data Visualization - Matplotlib/.ipynb_checkpoints/8.2 Matplotlib Pt 3-checkpoint.ipynb | CommunityOfCoders/ML_Workshop_Teachers |
Linewidth and Line Style | fig = plt.figure()
ax = fig.add_axes([0,0,1,1])
ax.plot(x,y,color='#008080') # Default Line width
# 5 times the linewidth
fig = plt.figure()
ax = fig.add_axes([0,0,1,1])
ax.plot(x,y,color='#008080',lw=5)#A shorthand is used here for linewidth which is lw
# To get transparency on the plotted line we can pass the alpha ... | _____no_output_____ | BSD-3-Clause | 08. Data Visualization - Matplotlib/.ipynb_checkpoints/8.2 Matplotlib Pt 3-checkpoint.ipynb | CommunityOfCoders/ML_Workshop_Teachers |
Markers* Markers are used when we have just a few number of data points. | # Say we have x an array of len(x) data points.
x
len(x)
# Say if we wanted to mark where those 11 points occured on the plot.
fig = plt.figure()
ax = fig.add_axes([0,0,1,1])
ax.plot(x,y,color='#008080',lw=3,marker='o',markersize=15,markerfacecolor='yellow',
markeredgewidth=3,markeredgecolor='black')
| _____no_output_____ | BSD-3-Clause | 08. Data Visualization - Matplotlib/.ipynb_checkpoints/8.2 Matplotlib Pt 3-checkpoint.ipynb | CommunityOfCoders/ML_Workshop_Teachers |
More examples on line and marker styles | fig, ax = plt.subplots(figsize=(12,6))
ax.plot(x, x+1, color="red", linewidth=0.25)
ax.plot(x, x+2, color="red", linewidth=0.50)
ax.plot(x, x+3, color="red", linewidth=1.00)
ax.plot(x, x+4, color="red", linewidth=2.00)
# possible linestype options ‘-‘, ‘–’, ‘-.’, ‘:’, ‘steps’
ax.plot(x, x+5, color="green", lw=3, line... | _____no_output_____ | BSD-3-Clause | 08. Data Visualization - Matplotlib/.ipynb_checkpoints/8.2 Matplotlib Pt 3-checkpoint.ipynb | CommunityOfCoders/ML_Workshop_Teachers |
Control over axis appearance* In this section we will look at controlling axis sizing properties in a matplotlib figure. | # Say we wanted to show the plot between 0 and 1 on the x-axis
fig = plt.figure()
ax = fig.add_axes([0,0,1,1])
ax.plot(x,y,color='#008080',lw=3,ls='--')
# Say we wanted to show the plot between 0 and 1 on the x-axis
fig = plt.figure()
ax = fig.add_axes([0,0,1,1])
ax.plot(x,y,color='#008080',lw=3,ls='--',
marker... | _____no_output_____ | BSD-3-Clause | 08. Data Visualization - Matplotlib/.ipynb_checkpoints/8.2 Matplotlib Pt 3-checkpoint.ipynb | CommunityOfCoders/ML_Workshop_Teachers |
Plot Range* We can configure the ranges of the axes using the set_ylim and set_xlim methods in the axis object, or axis('tight') for automatically getting \"tightly fitted\" axes ranges: | fig, axes = plt.subplots(1, 3, figsize=(12, 4))
axes[0].plot(x, x**2, label = 'X squaraed',color='red')
axes[0].plot(x, x**3,label='X cube',color='green')
axes[0].set_title("default axes ranges")
axes[1].plot(x, x**2, label = 'X squaraed',color='red')
axes[1].plot(x, x**3,label='X cube',color='green')
axes[1].axis('... | _____no_output_____ | BSD-3-Clause | 08. Data Visualization - Matplotlib/.ipynb_checkpoints/8.2 Matplotlib Pt 3-checkpoint.ipynb | CommunityOfCoders/ML_Workshop_Teachers |
Special Plot Types* There are many specialized plots we can create, such as barplots, histograms, scatter plots, and much more. Most of these type of plots we will actually create using seaborn, a statistical plotting library for Python. But here are a few examples of these type of plots: | # Scatter plot
plt.scatter(x,y)
# Histrogram
from random import sample
data = sample(range(1, 1000), 100)
plt.hist(data)
data = [np.random.normal(0, std, 100) for std in range(1, 4)]
# rectangular box plot
plt.boxplot(data,vert=True,patch_artist=True); | _____no_output_____ | BSD-3-Clause | 08. Data Visualization - Matplotlib/.ipynb_checkpoints/8.2 Matplotlib Pt 3-checkpoint.ipynb | CommunityOfCoders/ML_Workshop_Teachers |
--- 2. Select subsets from our dataset--- | from digits.data import matimport
from digits.data import select
dataroot='../../data/thomas/artcorr/'
imp = matimport.Importer(dataroot=dataroot) | _____no_output_____ | MIT | data/Selecting.ipynb | eegdigits/notebooks |
With `imp.open()` we can use HDF5 references to our samples and targets datasets without using up initial memory. The `samples` and `targets` objects are attached to the `store` attribute.In this notebook we will load the samples and targets from the file right away. | imp.open('3131.h5')
samples = imp.store.samples
targets = imp.store.targets
670*16
print(select.getsessionnames(samples))
for sess in select.getsessionnames(samples):
print(samples.xs(sess, level='session').shape[0]) | ['01', '07', '08', '09', '10', '11', '12', '13', '14', '15', '16']
632
650
652
652
683
687
669
658
610
672
609
| MIT | data/Selecting.ipynb | eegdigits/notebooks |
The functions in `digits.data.select` will provide a high level abstraction for subselecting and pruning the large dataset, specific to the studies parameters. For instance: column-wise+ select only sampling points from a time window with `select.fromtimerange(samples, min, max)`+ select all sampling points from a name... | print(select.getchannelnames(samples))
print(select.getsessionnames(samples))
print(select.getpresentationnames(samples))
print(select.getsessionnames(samples)) | ['01', '07', '08', '09', '10', '11', '12', '13', '14', '15', '16']
| MIT | data/Selecting.ipynb | eegdigits/notebooks |
The level/index names can be display with `head()` quite nicely: | samples.head() | _____no_output_____ | MIT | data/Selecting.ipynb | eegdigits/notebooks |
Now for the selection: | print(samples.shape)
print(select.getsessionnames(samples))
samples, targets = select.fromsessionlist(samples, targets, ['14', '15'])
samples.shape
samples = select.fromchannellist(samples, ['C1', 'C2'])
print(samples.shape)
samples = select.fromtimerange(samples, 't_0200', 't_0201')
print(samples.shape)
samples, targe... | \begin{tabular}{llllrrrr}
\toprule
& & & & C1 & & C2 & \\
& & & & t\_0200 & t\_0201 & t\_0200 & t\_0201 \\
subject & session & trial & presentation & & & & \\
\midrule
3131 & 14 & 2 & 1 & -7.291202... | MIT | data/Selecting.ipynb | eegdigits/notebooks |
DAT210x - Programming with Python for DS Module5- Lab3 | import pandas as pd
from datetime import timedelta
import matplotlib.pyplot as plt
import matplotlib
matplotlib.style.use('ggplot') # Look Pretty | _____no_output_____ | MIT | Module5/Module5 - Lab3.ipynb | azharmgh/pyrepo |
A convenience function for you to use: | def clusterInfo(model):
print("Cluster Analysis Inertia: ", model.inertia_)
print('------------------------------------------')
for i in range(len(model.cluster_centers_)):
print("\n Cluster ", i)
print(" Centroid ", model.cluster_centers_[i])
print(" #Samples ", (model.l... | _____no_output_____ | MIT | Module5/Module5 - Lab3.ipynb | azharmgh/pyrepo |
CDRs A [call detail record](https://en.wikipedia.org/wiki/Call_detail_record) (CDR) is a data record produced by a telephone exchange or other telecommunications equipment that documents the details of a telephone call or other telecommunications transaction (e.g., text message) that passes through that facility or de... | df1 = pd.read_csv('Datasets/CDR.csv')
df1 = df1.dropna()
df1['CallDate'] = pd.to_datetime(df1['CallDate'], 'coerce')
df1['CallTime'] = pd.to_timedelta(df1['CallTime'])
df1['Duration'] = pd.to_timedelta(df1['Duration'])
df1.dtypes | _____no_output_____ | MIT | Module5/Module5 - Lab3.ipynb | azharmgh/pyrepo |
Create a unique list of the phone number values (people) stored in the `In` column of the dataset, and save them in a regular python list called `unique_numbers`. Manually check through `unique_numbers` to ensure the order the numbers appear is the same order they (uniquely) appear in your dataset: | # .. your code here ..
unique_numbers = df1.In.unique().tolist()
unique_numbers | _____no_output_____ | MIT | Module5/Module5 - Lab3.ipynb | azharmgh/pyrepo |
Using some domain expertise, your intuition should direct you to know that people are likely to behave differently on weekends vs on weekdays: On Weekends1. People probably don't go into work1. They probably sleep in late on Saturday1. They probably run a bunch of random errands, since they couldn't during the week1. T... | print("Examining person: ", unique_numbers[0]) | Examining person: 4638472273
| MIT | Module5/Module5 - Lab3.ipynb | azharmgh/pyrepo |
Create a slice called `user1` that filters to only include dataset records where the `In` feature (user phone number) is equal to the first number on your unique list above: | # .. your code here ..
user1 = df1[df1['In'] == unique_numbers[0]]
user1 | _____no_output_____ | MIT | Module5/Module5 - Lab3.ipynb | azharmgh/pyrepo |
Alter your slice so that it includes only Weekday (Mon-Fri) values: | # .. your code here ..
pm5 = pd.to_timedelta('17:00:00')
am730 = pd.to_timedelta('07:30:00')
#user2 = user1[(((user1['DOW'] == 'Sat') | (user1['DOW'] == 'Sun')) & ((user1['CallTime'] > am1) & (user1['CallTime'] < am4)))]
user2 = user1
user1 = user1[(((user1['DOW'] == 'Mon') | (user1['DOW'] == 'Tue') | (user1['DOW'] == ... | _____no_output_____ | MIT | Module5/Module5 - Lab3.ipynb | azharmgh/pyrepo |
The idea is that the call was placed before 5pm. From Midnight-730a, the user is probably sleeping and won't call / wake up to take a call. There should be a brief time in the morning during their commute to work, then they'll spend the entire day at work. So the assumption is that most of the time is spent either at w... | # .. your code here .. | _____no_output_____ | MIT | Module5/Module5 - Lab3.ipynb | azharmgh/pyrepo |
Plot the Cell Towers the user connected to | # .. your code here ..
%matplotlib notebook
fig = plt.figure()
ax = fig.add_subplot(111)
ax.scatter(user1.TowerLon,user1.TowerLat, c='g', marker='o', alpha=0.2)
ax.set_title('Weedkay Calls (7:30am - 5pm)')
plt.show()
from sklearn.cluster import KMeans
def doKMeans(data, num_clusters=0):
# TODO: Be sure to only fe... | _____no_output_____ | MIT | Module5/Module5 - Lab3.ipynb | azharmgh/pyrepo |
Let's tun K-Means with `K=3` or `K=4`. There really should only be a two areas of concentration. If you notice multiple areas that are "hot" (multiple areas the user spends a lot of time at that are FAR apart from one another), then increase K=5, with the goal being that all centroids except two will sweep up the annoy... | model = doKMeans(user1, 4) | [[ 32.84579692 -96.81976265]
[ 32.89970164 -96.91026779]
[ 32.87348968 -96.85115015]
[ 32.911583 -96.892222 ]]
| MIT | Module5/Module5 - Lab3.ipynb | azharmgh/pyrepo |
Print out the mean `CallTime` value for the samples belonging to the cluster with the LEAST samples attached to it. If our logic is correct, the cluster with the MOST samples will be work. The cluster with the 2nd most samples will be home. And the `K=3` cluster with the least samples should be somewhere in between the... | midWayClusterIndices = clusterWithFewestSamples(model)
midWaySamples = user1[midWayClusterIndices]
print(" Its Waypoint Time: ", midWaySamples.CallTime.mean()) |
Cluster With Fewest Samples: 3
Its Waypoint Time: 0 days 07:44:31.892341
| MIT | Module5/Module5 - Lab3.ipynb | azharmgh/pyrepo |
Let's visualize the results! First draw the X's for the clusters: | fig1 = plt.figure()
ax1 = fig1.add_subplot(111)
ax1.scatter(model.cluster_centers_[:,1], model.cluster_centers_[:,0], s=169, c='r', marker='x', alpha=0.8, linewidths=2)
ax1.set_title('Weekday Calls Centroids')
plt.show()
clusterInfo(model)
users_phones = [2068627935,2894365987,1559410755,3688089071]
def examineNumber(... | Examining person: 2894365987
Cluster Analysis Inertia: 0.00584613804294
------------------------------------------
Cluster 0
Centroid [ 32.717667 -96.875194]
#Samples 141
Cluster 1
Centroid [ 32.72174109 -96.89194104]
#Samples 2705
Cluster 2
Centroid [ 32.741889 -96.857611]
#S... | MIT | Module5/Module5 - Lab3.ipynb | azharmgh/pyrepo |
2894365987 is the closest so far | examineNumber(df1,users_phones[2],4)
examineNumber(df1,users_phones[3],4)
def getClusterSamples(df, number, num_clusters):
print("getting cluster for person: ", number)
user = df[df['In'] == number]
pm5 = pd.to_timedelta('17:00:00')
am730 = pd.to_timedelta('07:30:00')
user = user[(((user['DOW'] == '... | examining user : 4638472273
getting cluster for person: 4638472273
Cluster With Fewest Samples: 1
Avg time : 0 days 07:44:01.395089
examining user : 1559410755
getting cluster for person: 1559410755
Cluster With Fewest Samples: 0
Avg time : 0 days 07:49:46.609049
examining user : 4931532174
getti... | MIT | Module5/Module5 - Lab3.ipynb | azharmgh/pyrepo |
Yapay Öğrenmeye Giriş IAli Taylan Cemgil Parametrik Regresyon, Parametrik Fonksyon Oturtma Problemi (Parametric Regression, Function Fitting)Verilen girdi ve çıktı ikilileri $x, y$ için parametrik bir fonksyon $f$ oturtma problemi. Parametre $w$ değerlerini öyle bir seçelim ki $$y \approx f(x; w)$$$x$: Girdi (Input)$y... | import matplotlib.pyplot as plt
import numpy as np
%matplotlib inline
from __future__ import print_function
from ipywidgets import interact, interactive, fixed
import ipywidgets as widgets
import matplotlib.pylab as plt
from IPython.display import clear_output, display, HTML
x = np.array([8.0 , 6.1 , 11., 7., 9., ... | _____no_output_____ | MIT | matkoy2021-1.ipynb | atcemgil/notes |
Rasgele Arama | x = np.array([8.0 , 6.1 , 11., 7., 9., 12. , 4., 2., 10, 5, 3])
y = np.array([6.04, 4.95, 5.58, 6.81, 6.33, 7.96, 5.24, 2.26, 8.84, 2.82, 3.68])
def hata(y, x, w):
N = len(y)
f = x*w[1]+w[0]
e = y-f
return np.sum(e*e)/2
w = np.array([0, 0])
E = hata(y, x, w)
for e in range(1000):
... | 999 6.88573142353
| MIT | matkoy2021-1.ipynb | atcemgil/notes |
Gerçek veri: Türkiyedeki araç sayıları | %matplotlib inline
import scipy as sc
import numpy as np
import pandas as pd
import matplotlib as mpl
import matplotlib.pylab as plt
df_arac = pd.read_csv(u'data/arac.csv',sep=';')
df_arac[['Year','Car']]
#df_arac
BaseYear = 1995
x = np.matrix(df_arac.Year[0:]).T-BaseYear
y = np.matrix(df_arac.Car[0:]).T/1000000.
pl... | _____no_output_____ | MIT | matkoy2021-1.ipynb | atcemgil/notes |
Örnek 1, devam: Modeli Öğrenmek* Öğrenmek: parametre kestirimi $w = [w_0, w_1]$* Genelde model veriyi hatasız açıklayamayacağı için her veri noktası için bir hata tanımlıyoruz:$$e_i = y_i - f(x_i; w)$$* Toplam kare hata $$E(w) = \frac{1}{2} \sum_i (y_i - f(x_i; w))^2 = \frac{1}{2} \sum_i e_i^2$$* Toplam kare hatayı $w... | from itertools import product
BaseYear = 1995
x = np.matrix(df_arac.Year[0:]).T-BaseYear
y = np.matrix(df_arac.Car[0:]).T/1000000.
# Setup the vandermonde matrix
N = len(x)
A = np.hstack((np.ones((N,1)), x))
left = -5
right = 15
bottom = -4
top = 6
step = 0.05
W0 = np.arange(left,right, step)
W1 = np.arange(bottom,t... | _____no_output_____ | MIT | matkoy2021-1.ipynb | atcemgil/notes |
Modeli Nasıl Kestirebiliriz? Fikir: En küçük kare hata (Gauss 1795, Legendre 1805)* Toplam hatanın $w_0$ ve $w_1$'e göre türevini hesapla, sıfıra eşitle ve çıkan denklemleri çöz\begin{eqnarray}\left(\begin{array}{c}y_0 \\ y_1 \\ \vdots \\ y_{N-1} \end{array}\right)\approx\left(\begin{array}{cc}1 & x_0 \\ 1 & x_1 \\ \v... | # Solving the Normal Equations
# Setup the Design matrix
N = len(x)
A = np.hstack((np.ones((N,1)), x))
#plt.imshow(A, interpolation='nearest')
# Solve the least squares problem
w_ls,E,rank,sigma = np.linalg.lstsq(A, y)
print('Parametreler: \nw0 = ', w_ls[0],'\nw1 = ', w_ls[1] )
print('Toplam Kare Hata:', E/2)
f = n... | Parametreler:
w0 = [[ 4.13258253]]
w1 = [[ 0.20987778]]
Toplam Kare Hata: [[ 37.19722385]]
| MIT | matkoy2021-1.ipynb | atcemgil/notes |
Polinomlar Parabol\begin{eqnarray}\left(\begin{array}{c}y_0 \\ y_1 \\ \vdots \\ y_{N-1} \end{array}\right)\approx\left(\begin{array}{ccc}1 & x_0 & x_0^2 \\ 1 & x_1 & x_1^2 \\ \vdots \\ 1 & x_{N-1} & x_{N-1}^2 \end{array}\right) \left(\begin{array}{c} w_0 \\ w_1 \\ w_2\end{array}\right)\end{eqnarray} $K$ derecesind... | x = np.array([10, 8, 13, 9, 11, 14, 6, 4, 12, 7, 5])
N = len(x)
x = x.reshape((N,1))
y = np.array([8.04, 6.95, 7.58, 8.81, 8.33, 9.96, 7.24, 4.26, 10.84, 4.82, 5.68]).reshape((N,1))
#y = np.array([9.14, 8.14, 8.74, 8.77, 9.26, 8.10, 6.13, 3.10, 9.13, 7.26, 4.74]).reshape((N,1))
#y = np.array([7.46, 6.77, 12.74, 7.11, 7... | _____no_output_____ | MIT | matkoy2021-1.ipynb | atcemgil/notes |
About https://www.kaggle.com/uladzimirkapeika/feature-engineering-lightgbm-top-1https://zhuanlan.zhihu.com/p/145969470 Version 1.0 Libraries> Check your versions | import pandas as pd
import numpy as np
from itertools import product
import sklearn
from sklearn.preprocessing import LabelEncoder
from sklearn.linear_model import LinearRegression
from sklearn.neighbors import KNeighborsRegressor
from sklearn.ensemble import RandomForestRegressor
import lightgbm as lgb
import calendar... | numpy 1.20.1
pandas 1.2.3
seaborn 0.11.1
sklearn 0.24.1
lightgbm 3.2.0
| MIT | {{cookiecutter.project_slug}}/notebooks/02_sj_salse_predict_ml.ipynb | juforg/cookiecutter-ds-py3tkinter |
**Load data** | test = pd.read_csv('../data/0_raw/sales/test.csv.gz')
sales = pd.read_csv('../data/0_raw/sales/sales_train.csv.gz', encoding='UTF-8')
# sales = pd.read_csv('https://storage.googleapis.com/kaggle-competitions-data/kaggle-v2/8587/868304/compressed/sales_train.csv.zip?GoogleAccessId=web-data@kaggle-161607.iam.gserviceacco... | 214200
2935849
60
22170
84
item_name item_id \
0 ! ВО ВЛАСТИ НАВАЖДЕНИЯ (ПЛАСТ.) D 0
1 !ABBYY FineReader 12 Professional Edition Full... 1
2 ***В ЛУЧАХ СЛАВЫ (UNV) D 2
3 ***ГОЛУБАЯ ВОЛНА (Univ) ... | MIT | {{cookiecutter.project_slug}}/notebooks/02_sj_salse_predict_ml.ipynb | juforg/cookiecutter-ds-py3tkinter |
**Create dataset** Remove outliers | sns.boxplot(x=sales.item_cnt_day)
sns.boxplot(x=sales.item_price)
train = sales[(sales.item_price < 100000) & (sales.item_price > 0)]
train = train[sales.item_cnt_day < 1001] | /Users/songjie/.local/share/virtualenvs/snp_mvp-8ex0mMfN/lib/python3.7/site-packages/ipykernel_launcher.py:2: UserWarning: Boolean Series key will be reindexed to match DataFrame index.
| MIT | {{cookiecutter.project_slug}}/notebooks/02_sj_salse_predict_ml.ipynb | juforg/cookiecutter-ds-py3tkinter |
Detect same shops | print(shops[shops.shop_id.isin([0, 57])]['shop_name'])
print(shops[shops.shop_id.isin([1, 58])]['shop_name'])
print(shops[shops.shop_id.isin([40, 39])]['shop_name'])
train.loc[train.shop_id == 0, 'shop_id'] = 57
test.loc[test.shop_id == 0, 'shop_id'] = 57
train.loc[train.shop_id == 1, 'shop_id'] = 58
test.loc[test.sho... | _____no_output_____ | MIT | {{cookiecutter.project_slug}}/notebooks/02_sj_salse_predict_ml.ipynb | juforg/cookiecutter-ds-py3tkinter |
Simple train dataset | index_cols = ['shop_id', 'item_id', 'date_block_num']
df = []
for block_num in train['date_block_num'].unique():
cur_shops = train.loc[sales['date_block_num'] == block_num, 'shop_id'].unique()
cur_items = train.loc[sales['date_block_num'] == block_num, 'item_id'].unique()
df.append(np.array(list(product(*... | _____no_output_____ | MIT | {{cookiecutter.project_slug}}/notebooks/02_sj_salse_predict_ml.ipynb | juforg/cookiecutter-ds-py3tkinter |
Add test | test['date_block_num'] = 34
test['date_block_num'] = test['date_block_num'].astype(np.int8)
test['shop_id'] = test['shop_id'].astype(np.int8)
test['item_id'] = test['item_id'].astype(np.int16)
df = pd.concat([df, test], ignore_index=True, sort=False, keys=index_cols)
df.fillna(0, inplace=True) | _____no_output_____ | MIT | {{cookiecutter.project_slug}}/notebooks/02_sj_salse_predict_ml.ipynb | juforg/cookiecutter-ds-py3tkinter |
Feature engineering **Shop features*** City of a shop* City coords* Country part (0-4) based on the map | shops['city'] = shops['shop_name'].apply(lambda x: x.split()[0].lower())
shops.loc[shops.city == '!якутск', 'city'] = 'якутск'
shops['city_code'] = LabelEncoder().fit_transform(shops['city'])
coords = dict()
coords['якутск'] = (62.028098, 129.732555, 4)
coords['адыгея'] = (44.609764, 40.100516, 3)
coords['балашиха'] =... | _____no_output_____ | MIT | {{cookiecutter.project_slug}}/notebooks/02_sj_salse_predict_ml.ipynb | juforg/cookiecutter-ds-py3tkinter |
**Item features*** Item category* More common item category | map_dict = {
'Чистые носители (штучные)': 'Чистые носители',
'Чистые носители (шпиль)' : 'Чистые носители',
'PC ': 'Аксессуары',
'Служебные': 'Служебные '
}
items = pd.merge(items, item_cats, on='item_category_id')
items['item_category'] = items['item_catego... | _____no_output_____ | MIT | {{cookiecutter.project_slug}}/notebooks/02_sj_salse_predict_ml.ipynb | juforg/cookiecutter-ds-py3tkinter |
**Date features*** Weekends count (4 or 5)* Number of days in a month | def count_days(date_block_num):
year = 2013 + date_block_num // 12
month = 1 + date_block_num % 12
weeknd_count = len([1 for i in calendar.monthcalendar(year, month) if i[6] != 0])
days_in_month = calendar.monthrange(year, month)[1]
return weeknd_count, days_in_month, month
map_dict = {i: count_day... | _____no_output_____ | MIT | {{cookiecutter.project_slug}}/notebooks/02_sj_salse_predict_ml.ipynb | juforg/cookiecutter-ds-py3tkinter |
**Interaction features*** Item is new* Item was bought in this shop before | first_item_block = df.groupby(['item_id'])['date_block_num'].min().reset_index()
first_item_block['item_first_interaction'] = 1
first_shop_item_buy_block = df[df['date_block_num'] > 0].groupby(['shop_id', 'item_id'])['date_block_num'].min().reset_index()
first_shop_item_buy_block['first_date_block_num'] = first_shop_i... | _____no_output_____ | MIT | {{cookiecutter.project_slug}}/notebooks/02_sj_salse_predict_ml.ipynb | juforg/cookiecutter-ds-py3tkinter |
**Basic lag features** | def lag_feature(df, lags, col):
tmp = df[['date_block_num','shop_id','item_id',col]]
for i in lags:
shifted = tmp.copy()
shifted.columns = ['date_block_num','shop_id','item_id', col+'_lag_'+str(i)]
shifted['date_block_num'] += i
df = pd.merge(df, shifted, on=['date_block_num','sh... | _____no_output_____ | MIT | {{cookiecutter.project_slug}}/notebooks/02_sj_salse_predict_ml.ipynb | juforg/cookiecutter-ds-py3tkinter |
**Target encoding** | #Add target encoding for items for last 3 months
item_id_target_mean = df.groupby(['date_block_num','item_id'])['item_cnt_month'].mean().reset_index().rename(columns={"item_cnt_month": "item_target_enc"}, errors="raise")
df = pd.merge(df, item_id_target_mean, on=['date_block_num','item_id'], how='left')
df['item_targ... | _____no_output_____ | MIT | {{cookiecutter.project_slug}}/notebooks/02_sj_salse_predict_ml.ipynb | juforg/cookiecutter-ds-py3tkinter |
Extra interaction features | #For new items add avg category sales for last 3 months
item_id_target_mean = df[df['item_first_interaction'] == 1].groupby(['date_block_num','item_category_code'])['item_cnt_month'].mean().reset_index().rename(columns={
"item_cnt_month": "new_item_cat_avg"}, errors="raise")
df = pd.merge(df, item_id_target_mean, ... | _____no_output_____ | MIT | {{cookiecutter.project_slug}}/notebooks/02_sj_salse_predict_ml.ipynb | juforg/cookiecutter-ds-py3tkinter |
Add sales for the last three months for similar item (item with id = item_id - 1;kinda tricky feature, but increased the metric significantly) | def lag_feature_adv(df, lags, col):
tmp = df[['date_block_num','shop_id','item_id',col]]
for i in lags:
shifted = tmp.copy()
shifted.columns = ['date_block_num','shop_id','item_id', col+'_lag_'+str(i)+'_adv']
shifted['date_block_num'] += i
shifted['item_id'] -= 1
df = pd.... | _____no_output_____ | MIT | {{cookiecutter.project_slug}}/notebooks/02_sj_salse_predict_ml.ipynb | juforg/cookiecutter-ds-py3tkinter |
Remove data for the first three months | df.fillna(0, inplace=True)
df = df[(df['date_block_num'] > 2)]
df.head()
df.columns
#Save dataset
df.drop(['ID'], axis=1, inplace=True, errors='ignore')
df.to_pickle('../output/models/sales_df.pkl') | _____no_output_____ | MIT | {{cookiecutter.project_slug}}/notebooks/02_sj_salse_predict_ml.ipynb | juforg/cookiecutter-ds-py3tkinter |
Train model | df = pd.read_pickle('../output/models/sales_df.pkl')
df.info()
X_train = df[df.date_block_num < 33].drop(['item_cnt_month'], axis=1)
Y_train = df[df.date_block_num < 33]['item_cnt_month']
X_valid = df[df.date_block_num == 33].drop(['item_cnt_month'], axis=1)
Y_valid = df[df.date_block_num == 33]['item_cnt_month']
X_tes... | _____no_output_____ | MIT | {{cookiecutter.project_slug}}/notebooks/02_sj_salse_predict_ml.ipynb | juforg/cookiecutter-ds-py3tkinter |
Stacking didn't work for me. I'd tried 2 approaches:1. XGBoost + CatBoost + LightGBM at the first level and LinearRegression/LightGBM at the second level1. LinearRegression + LightGBM + RandomForest at the first level and LinearRegression/LightGBM at the second level | test = pd.read_csv('../data/0_raw/sales/test.csv.gz')
Y_test = gbm.predict(X_test[feature_name]).clip(0, 20)
submission = pd.DataFrame({
"ID": test.index,
"item_cnt_month": Y_test
})
submission.to_csv('../output/gbm_submission.csv', index=False) | _____no_output_____ | MIT | {{cookiecutter.project_slug}}/notebooks/02_sj_salse_predict_ml.ipynb | juforg/cookiecutter-ds-py3tkinter |
Displaying Surfacespy3Dmol supports the following surface types:* VDW - van der Waals surface* MS - molecular surface* SES - solvent excluded surface* SAS - solvent accessible surface | import py3Dmol | _____no_output_____ | Apache-2.0 | 1-3D-visualization/4-Surfaces.ipynb | NicholasAKovacs/mmtf-workshop |
Add surfaceIn the structure below (HLA complex with antigen peptide pVR), we add a solvent excluded surface (SES) to the heavy chain to highlight the binding pocket for the antigen peptide (rendered as spheres). | viewer = py3Dmol.view(query='pdb:5XS3')
heavychain = {'chain':'A'}
lightchain = {'chain':'B'}
antigen = {'chain':'C'}
viewer.setStyle(heavychain,{'cartoon':{'color':'blue'}})
viewer.setStyle(lightchain,{'cartoon':{'color':'yellow'}})
viewer.setStyle(antigen,{'sphere':{'colorscheme':'orangeCarbon'}})
viewer.addSurfac... | _____no_output_____ | Apache-2.0 | 1-3D-visualization/4-Surfaces.ipynb | NicholasAKovacs/mmtf-workshop |
NETWORK = "https://nvlabs-fi-cdn.nvidia.com/stylegan2-ada-pytorch/pretrained/ffhq.pkl"
STEPS = 300
FPS = 30
FREEZE_STEPS = 30 | _____no_output_____ | MIT | StyleGAN2.ipynb | patprem/FaceMorphing | |
Upload Starting ImageChoose your starting image. | import os
from google.colab import files
uploaded = files.upload()
if len(uploaded) != 1:
print("Upload exactly 1 file for source.")
else:
for k, v in uploaded.items():
_, ext = os.path.splitext(k)
os.remove(k)
SOURCE_NAME = f"source{ext}"
open(SOURCE_NAME, 'wb').write(v) | _____no_output_____ | MIT | StyleGAN2.ipynb | patprem/FaceMorphing |
Also, choose your ending image. | uploaded = files.upload()
if len(uploaded) != 1:
print("Upload exactly 1 file for target.")
else:
for k, v in uploaded.items():
_, ext = os.path.splitext(k)
os.remove(k)
TARGET_NAME = f"target{ext}"
open(TARGET_NAME, 'wb').write(v) | _____no_output_____ | MIT | StyleGAN2.ipynb | patprem/FaceMorphing |
Install SoftwareSome software must be installed into Colab, for this notebook to work. We are specificially using these technologies:* [Training Generative Adversarial Networks with Limited Data](https://arxiv.org/abs/2006.06676)Tero Karras, Miika Aittala, Janne Hellsten, Samuli Laine, Jaakko Lehtinen, Timo Aila* [One... | !wget http://dlib.net/files/shape_predictor_5_face_landmarks.dat.bz2
!bzip2 -d shape_predictor_5_face_landmarks.dat.bz2
import sys
!git clone https://github.com/NVlabs/stylegan2-ada-pytorch.git
!pip install ninja
sys.path.insert(0, "/content/stylegan2-ada-pytorch") | fatal: destination path 'stylegan2-ada-pytorch' already exists and is not an empty directory.
Requirement already satisfied: ninja in /usr/local/lib/python3.7/dist-packages (1.10.2.3)
| MIT | StyleGAN2.ipynb | patprem/FaceMorphing |
Preprocess Images for Best StyleGAN ResultsThe following are helper functions for the preprocessing. | import cv2
import numpy as np
from PIL import Image
import dlib
detector = dlib.get_frontal_face_detector()
predictor = dlib.shape_predictor('shape_predictor_5_face_landmarks.dat')
def find_eyes(img):
gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
rects = detector(gray, 0)
if len(rects) == 0:
raise ValueEr... | _____no_output_____ | MIT | StyleGAN2.ipynb | patprem/FaceMorphing |
The following will preprocess and crop your images. If you receive an error indicating multiple faces were found, try to crop your image better or obscure the background. If the program does not see a face, then attempt to obtain a clearer and more high-resolution image. | from matplotlib import pyplot as plt
import cv2
image_source = cv2.imread(SOURCE_NAME)
if image_source is None:
raise ValueError("Source image not found")
image_target = cv2.imread(TARGET_NAME)
if image_target is None:
raise ValueError("Source image not found")
cropped_source = crop_stylegan(image_source)
cr... | _____no_output_____ | MIT | StyleGAN2.ipynb | patprem/FaceMorphing |
Convert Source to a GANFirst, we convert the source to a GAN latent vector. This process will take several minutes. | cmd = f"python /content/stylegan2-ada-pytorch/projector.py --save-video 0 --num-steps 1000 --outdir=out_source --target=cropped_source.png --network={NETWORK}"
!{cmd} | Loading networks from "https://nvlabs-fi-cdn.nvidia.com/stylegan2-ada-pytorch/pretrained/ffhq.pkl"...
Computing W midpoint and stddev using 10000 samples...
Setting up PyTorch plugin "bias_act_plugin"... Done.
Setting up PyTorch plugin "upfirdn2d_plugin"... Done.
step 1/1000: dist 0.64 loss 24569.53
step 2/1000: ... | MIT | StyleGAN2.ipynb | patprem/FaceMorphing |
Convert Target to a GANNext, we convert the target to a GAN latent vector. This process will also take several minutes. | cmd = f"python /content/stylegan2-ada-pytorch/projector.py --save-video 0 --num-steps 1000 --outdir=out_target --target=cropped_target.png --network={NETWORK}"
!{cmd} | Loading networks from "https://nvlabs-fi-cdn.nvidia.com/stylegan2-ada-pytorch/pretrained/ffhq.pkl"...
Computing W midpoint and stddev using 10000 samples...
Setting up PyTorch plugin "bias_act_plugin"... Done.
Setting up PyTorch plugin "upfirdn2d_plugin"... Done.
step 1/1000: dist 0.55 loss 24569.43
step 2/1000: ... | MIT | StyleGAN2.ipynb | patprem/FaceMorphing |
With the conversion complete, lets have a look at the two GANs. | img_gan_source = cv2.imread('/content/out_source/proj.png')
img = cv2.cvtColor(img_gan_source, cv2.COLOR_BGR2RGB)
plt.imshow(img)
plt.title('source-gan')
plt.show()
img_gan_target = cv2.imread('/content/out_target/proj.png')
img = cv2.cvtColor(img_gan_target, cv2.COLOR_BGR2RGB)
plt.imshow(img)
plt.title('target-gan')
p... | _____no_output_____ | MIT | StyleGAN2.ipynb | patprem/FaceMorphing |
Build the VideoThe following code builds a transition video between the two latent vectors previously obtained. | import torch
import dnnlib
import legacy
import PIL.Image
import numpy as np
import imageio
from tqdm.notebook import tqdm
lvec1 = np.load('/content/out_source/projected_w.npz')['w']
lvec2 = np.load('/content/out_target/projected_w.npz')['w']
network_pkl = "https://nvlabs-fi-cdn.nvidia.com/stylegan2-ada-pytorch/pretr... | _____no_output_____ | MIT | StyleGAN2.ipynb | patprem/FaceMorphing |
Download your VideoIf you made it through all of these steps, you are now ready to download your video. | from google.colab import files
files.download("movie.mp4") | _____no_output_____ | MIT | StyleGAN2.ipynb | patprem/FaceMorphing |
k-Nearest Neighbor (kNN) exercise*Complete and hand in this completed worksheet (including its outputs and any supporting code outside of the worksheet) with your assignment submission. For more details see the [assignments page](http://vision.stanford.edu/teaching/cs231n/assignments.html) on the course website.*The k... | # Run some setup code for this notebook.
import random
import numpy as np
from cs231n.data_utils import load_CIFAR10
import matplotlib.pyplot as plt
# This is a bit of magic to make matplotlib figures appear inline in the notebook
# rather than in a new window.
%matplotlib inline
plt.rcParams['figure.figsize'] = (10.... | _____no_output_____ | MIT | assignment1/knn.ipynb | meijun/cs231n-assignment |
We would now like to classify the test data with the kNN classifier. Recall that we can break down this process into two steps: 1. First we must compute the distances between all test examples and all train examples. 2. Given these distances, for each test example we find the k nearest examples and have them vote for t... | # Open cs231n/classifiers/k_nearest_neighbor.py and implement
# compute_distances_two_loops.
# Test your implementation:
dists = classifier.compute_distances_two_loops(X_test)
print dists.shape
# We can visualize the distance matrix: each row is a single test example and
# its distances to training examples
plt.imshow... | _____no_output_____ | MIT | assignment1/knn.ipynb | meijun/cs231n-assignment |
**Inline Question 1:** Notice the structured patterns in the distance matrix, where some rows or columns are visible brighter. (Note that with the default color scheme black indicates low distances while white indicates high distances.)- What in the data is the cause behind the distinctly bright rows?- What causes the ... | # Now implement the function predict_labels and run the code below:
# We use k = 1 (which is Nearest Neighbor).
y_test_pred = classifier.predict_labels(dists, k=1)
# Compute and print the fraction of correctly predicted examples
num_correct = np.sum(y_test_pred == y_test)
accuracy = float(num_correct) / num_test
print... | Got 137 / 500 correct => accuracy: 0.274000
| MIT | assignment1/knn.ipynb | meijun/cs231n-assignment |
You should expect to see approximately `27%` accuracy. Now lets try out a larger `k`, say `k = 5`: | y_test_pred = classifier.predict_labels(dists, k=5)
num_correct = np.sum(y_test_pred == y_test)
accuracy = float(num_correct) / num_test
print 'Got %d / %d correct => accuracy: %f' % (num_correct, num_test, accuracy) | Got 139 / 500 correct => accuracy: 0.278000
| MIT | assignment1/knn.ipynb | meijun/cs231n-assignment |
You should expect to see a slightly better performance than with `k = 1`. | # Now lets speed up distance matrix computation by using partial vectorization
# with one loop. Implement the function compute_distances_one_loop and run the
# code below:
dists_one = classifier.compute_distances_one_loop(X_test)
# To ensure that our vectorized implementation is correct, we make sure that it
# agrees ... | Two loop version took 25.608644 seconds
One loop version took 49.357512 seconds
No loop version took 0.393901 seconds
| MIT | assignment1/knn.ipynb | meijun/cs231n-assignment |
Cross-validationWe have implemented the k-Nearest Neighbor classifier but we set the value k = 5 arbitrarily. We will now determine the best value of this hyperparameter with cross-validation. | num_folds = 5
k_choices = [1, 3, 5, 8, 10, 12, 15, 20, 50, 100]
X_train_folds = []
y_train_folds = []
################################################################################
# TODO: #
# Split up the training data into folds. After splittin... | Got 141 / 500 correct => accuracy: 0.282000
| MIT | assignment1/knn.ipynb | meijun/cs231n-assignment |
Function call parameter rules and unpackingPython functions have the following four forms when declaring parameters:1. Without default value: `def func(a): pass`1. With default value: `def func(a, b = 1): pass`1. Arbitrary position parameters: `def func(a, b = 1, *c): pass`1. Arbitrary key parameter: `def func(a, b = ... | def func(a, b = 1):
pass
func(a = "G", 20) # SyntaxError | _____no_output_____ | MIT | Notebooks/Arguments-and-Unpacking.ipynb | gtavasoli/PyTips |
Another rule is position parameter priority: | def func(a, b = 1):
pass
func(20, a = "G") # TypeError: Repeat assignment to parameter a | _____no_output_____ | MIT | Notebooks/Arguments-and-Unpacking.ipynb | gtavasoli/PyTips |
**The safest way is to use all keyword parameters.** Arbitrary ParametersAny parameter can accept any number of parameters, where the form of `*a` represents any number of **positional parameters**, and `**d` represents any number of **keyword parameters**: | def concat(*lst, sep = "/"):
return sep.join((str(i) for i in lst))
print(concat("G", 20, "@", "Hz", sep = "")) | G20@Hz
| MIT | Notebooks/Arguments-and-Unpacking.ipynb | gtavasoli/PyTips |
The syntax of the above `def concat(*lst, sep = "/")` was proposed by [PEP 3102](https://www.python.org/dev/peps/pep-3102/) and implemented after **Python 3.0**. The keyword function here must be clearly specified and cannot be inferred by position: | print(concat("G", 20, "-")) # Not G-20 | G/20/-
| MIT | Notebooks/Arguments-and-Unpacking.ipynb | gtavasoli/PyTips |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.