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 |
|---|---|---|---|---|---|
Task: implement TF-IDF featuresNot all words are equally useful. One can prioritize rare words and downscale words like "and"/"or" by using __tf-idf features__. This abbreviation stands for __text frequency/inverse document frequence__ and means exactly that:$$ feature_i = { Count(word_i \in x) \times { log {N \over C... | def tf_idf(texts_collection, a = 1):
global_counts = text_to_bow(' '.join(texts_collection))
return np.stack(list(map(text_to_bow, texts_train))) * np.log(len(texts_collection)/(global_counts + a))
idf = np.log(len(texts_train)/(global_counts + 1))
X_train_tfidf = X_train_bow * idf
X_test_tfidf = X_test_bow * i... | _____no_output_____ | MIT | week02_classification/homework_part1.ipynb | andnlv/natural-language-processing |
```````````````````````````````````````````````` Solving it better: word vectorsLet's try another approach: instead of counting per-word frequencies, we shall map all words to pre-trained word vectors and average over them to get text features.This should give us two key advantages: (1) we now have 10^2 features instea... | import gensim.downloader
embeddings = gensim.downloader.load("fasttext-wiki-news-subwords-300")
# If you're low on RAM or download speed, use "glove-wiki-gigaword-100" instead. Ignore all further asserts.
type(embeddings)
embeddings.get_vector('qweqw', )
def vectorize_sum(comment):
"""
implement a function th... | _____no_output_____ | MIT | week02_classification/homework_part1.ipynb | andnlv/natural-language-processing |
Assignments of Day 6 | # Create a bank account class that has two attributes
# arr = []
# arr.append([ownername,balance])
# for i in arr:
# if self.name == i[0]:
class bank:
def __init__(self,ownername,balance):
self.ownername = ownername
self.balance = balance
def deposit(self,amount):
self.... | you don't have enough cradit in your account, see you have only 5000
| Apache-2.0 | Day6 Assignment.ipynb | divi2398/Lakshmi.M |
For this challenge,create a cone class that has two attributes:*R=Radius*h=HeightAnd two methods:*Volume = Π * r2 = (h/3)*Surface area : base : Π * r2 , side : Π * r * √(r2 + h2)Make only one class with functions,as in where required import Math. | import math
class cone:
def __init__(self,radius,height):
self.radius=radius
self.height=height
def volume(self):
vol = math.pi * (self.radius**2) * (self.height/3)
print("Volume of this cone is : ",vol)
def surfaceArea(self):
area = math.pi* self.ra... | Surface area of this cone is 201.22293136239685
| Apache-2.0 | Day6 Assignment.ipynb | divi2398/Lakshmi.M |
Data analysis with Python, Apache Spark, and PixieDust***In this notebook you will:* analyze customer demographics, such as, age, gender, income, and location* combine that data with sales data to examine trends for product categories, transaction types, and product popularity* load data from GitHub as well as from a... | import pixiedust
import pyspark.sql.functions as func
import pyspark.sql.types as types
import re
import json
import os
import requests | _____no_output_____ | Apache-2.0 | notebooks/analyze-customer-data.ipynb | aimazin/customer-data-with-spark |
**If you get any errors or if a package is out of date:*** uncomment the lines in the next cell (remove the ``)* restart the kernel (from the Kernel menu at the top of the notebook)* reload the browser page* run the cell above, and continue with the notebook | #!pip install jinja2 --user --upgrade
#!pip install pixiedust --user --upgrade
#!pip install -U --no-deps bokeh | _____no_output_____ | Apache-2.0 | notebooks/analyze-customer-data.ipynb | aimazin/customer-data-with-spark |
Load data into the notebookThe data file contains both the customer demographic data that you'll analyzed in Part 1, and the sales transaction data for Part 2.With `pixiedust.sampleData()` you can load csv data from any url. The below loads the data in a Spark DataFrame. > In case you wondered, this works with Pandas ... | raw_df = pixiedust.sampleData('https://raw.githubusercontent.com/IBM/analyze-customer-data-spark-pixiedust/master/data/customers_orders1_opt.csv')
raw_df | _____no_output_____ | Apache-2.0 | notebooks/analyze-customer-data.ipynb | aimazin/customer-data-with-spark |
[Back to Table of Contents](toc) Explore customer demographics In this part of the notebook, you will prepare the customer data and then start learning about your customers by creating multiple charts and maps. Prepare the customer data setCreate a new Spark DataFrame with only the data you need and then cleanse and ... | # Extract the customer information from the data set
customer_df = raw_df.select("CUST_ID",
"CUSTNAME",
"ADDRESS1",
"ADDRESS2",
"CITY",
"POSTAL_CODE",
... | _____no_output_____ | Apache-2.0 | notebooks/analyze-customer-data.ipynb | aimazin/customer-data-with-spark |
Notice that the data type of the AGE column is currently a string. Convert the AGE column to a numeric data type so you can run calculations on customer age. | # ---------------------------------------
# Cleanse age (enforce numeric data type)
# ---------------------------------------
def getNumericVal(col):
"""
input: pyspark.sql.types.Column
output: the numeric value represented by col or None
"""
try:
return int(col)
except ValueError:
... | _____no_output_____ | Apache-2.0 | notebooks/analyze-customer-data.ipynb | aimazin/customer-data-with-spark |
The GenderCode column contains salutations instead of gender values. Derive the gender information for each customer based on the salutation and rename the GenderCode column to GENDER. | # ------------------------------
# Derive gender from salutation
# ------------------------------
def deriveGender(col):
""" input: pyspark.sql.types.Column
output: "male", "female" or "unknown"
"""
if col in ['Mr.', 'Master.']:
return 'male'
elif col in ['Mrs.', 'Miss.']:
re... | _____no_output_____ | Apache-2.0 | notebooks/analyze-customer-data.ipynb | aimazin/customer-data-with-spark |
Explore the customer data setInstead of exploring the data with `.printSchema()` and `.show()` you can quickly explore data sets using PixieDust'. Invoke the `display()` command and click the table icon to review the schema and preview the data. Customize the options to display only a subset of the fields or rows or a... | display(customer_df) | _____no_output_____ | Apache-2.0 | notebooks/analyze-customer-data.ipynb | aimazin/customer-data-with-spark |
[Back to Table of Contents](toc) Visualize customer demographics and locationsNow you are ready to explore the customer base. Using simple charts, you can quickly see these characteristics: * Customer demographics (gender and age) * Customer locations (city, state, and country)You will create charts with the PixieDust ... | display(customer_df) | _____no_output_____ | Apache-2.0 | notebooks/analyze-customer-data.ipynb | aimazin/customer-data-with-spark |
[Back to Table of Contents](toc) View customers by generation in a bar chartLook at how many customers you have per "generation."Run the next cell and configure the graph: 1. Choose **Bar Chart** as the chart type and configure the chart options as instructed below.2. Put `GENERATION` into the **Keys** box.3. Set **agg... | display(customer_df) | _____no_output_____ | Apache-2.0 | notebooks/analyze-customer-data.ipynb | aimazin/customer-data-with-spark |
[Back to Table of Contents](toc) View customers by age in a histogram chartA generation is a broad age range. You can look at a smaller age range with a histogram chart. A histogram is like a bar chart except each bar represents a range of numbers, called a bin. You can customize the size of the age range by adjusting ... | display(customer_df) | _____no_output_____ | Apache-2.0 | notebooks/analyze-customer-data.ipynb | aimazin/customer-data-with-spark |
[Back to Table of Contents](toc) View specific information with a filter functionYou can filter records to restrict analysis by using the [PySpark DataFrame](https://spark.apache.org/docs/latest/api/python/pyspark.sql.htmlpyspark.sql.DataFrame) `filter()` function.If you want to view the age distribution for a specific... | # Data subsetting: display age distribution for a specific generation
# (Chart type: histogram, Chart Options > Values: AGE)
# to change the filter condition remove the # sign
condition = "GENERATION = 'Baby_Boomers'"
#condition = "GENERATION = 'Gen_X'"
#condition = "GENERATION = 'Gen_Y'"
#condition = "GENERATION = 'G... | _____no_output_____ | Apache-2.0 | notebooks/analyze-customer-data.ipynb | aimazin/customer-data-with-spark |
PixieDust supports basic filtering to make it easy to analyse data subsets. For example, to view the age distribution for a specific gender configure the chart as follows: 1. Choose `Histogram` as the chart type. 2. Put `AGE` into the **Values** box and click OK. 3. Click the filter button (looking like a funnel), a... | display(customer_df) | _____no_output_____ | Apache-2.0 | notebooks/analyze-customer-data.ipynb | aimazin/customer-data-with-spark |
You can also filter by location. For example, the following command creates a new DataFrame that filters for customers from the USA: | condition = "COUNTRY_CODE = 'US'"
us_customer_df = customer_df.filter(condition) | _____no_output_____ | Apache-2.0 | notebooks/analyze-customer-data.ipynb | aimazin/customer-data-with-spark |
You can pivot your analysis perspective based on aspects that are of interest to you by choosing different keys and clusters.Create a bar chart and cluster the data.Run the next cell and configure the graph:1. Choose **Bar chart** as the chart type.2. Put `COUNTRY_CODE` into the **Keys** box.4. Set Aggregation to **COU... | display(us_customer_df) | _____no_output_____ | Apache-2.0 | notebooks/analyze-customer-data.ipynb | aimazin/customer-data-with-spark |
Now try to cluster the customers by state.A bar chart isn't the best way to show geographic location! [Back to Table of Contents](toc) View customer density by location with a mapMaps are a much better way to view location data than other chart types. Visualize customer density by US state with a map.Run the next cell ... | display(us_customer_df) | _____no_output_____ | Apache-2.0 | notebooks/analyze-customer-data.ipynb | aimazin/customer-data-with-spark |
[Back to Table of Contents](toc) Enrich demographic information with open dataYou can easily combine other sources of data with your existing data. There is a lot of publicly available open data sets that can be very helpful. For example, knowing the approximate income level of your customers might help you target your... | # Load median income information for all US ZIP codes from a public source
income_df = pixiedust.sampleData('https://raw.githubusercontent.com/IBM/analyze-customer-data-spark-pixiedust/master/data/x19_income_select.csv')
income_df.printSchema() | _____no_output_____ | Apache-2.0 | notebooks/analyze-customer-data.ipynb | aimazin/customer-data-with-spark |
Now cleanse the income data set to remove the data that you don't need. Create a new DataFrame for this data: - The zip code, extracted from the GEOID column. - The column B19049e1, which contains the median household income for 2013. | # ------------------------------
# Helper: Extract ZIP code
# ------------------------------
def extractZIPCode(col):
""" input: pyspark.sql.types.Column containing a geo code, like '86000US01001'
output: ZIP code
"""
m = re.match('^\d+US(\d\d\d\d\d)$',col)
if m:
return m.group(1)
el... | _____no_output_____ | Apache-2.0 | notebooks/analyze-customer-data.ipynb | aimazin/customer-data-with-spark |
Perform a left outer join on the customer data set with the income data set, using the zip code as the join condition. For the complete syntax of joins, go to the pyspark DataFrame documentation and scroll down to the `join` syntax. | us_customer_df = us_customer_df.join(income_df, us_customer_df.POSTAL_CODE == income_df.ZIP, 'left_outer').drop('GEOID').drop('ZIP')
display(us_customer_df) | _____no_output_____ | Apache-2.0 | notebooks/analyze-customer-data.ipynb | aimazin/customer-data-with-spark |
REINFORCE in lasagneJust like we did before for q-learning, this time we'll design a lasagne network to learn `CartPole-v0` via policy gradient (REINFORCE).Most of the code in this notebook is taken from approximate qlearning, so you'll find it more or less familiar and even simpler. __Frameworks__ - we'll accept this... | %env THEANO_FLAGS = 'floatX=float32'
import os
if type(os.environ.get("DISPLAY")) is not str or len(os.environ.get("DISPLAY")) == 0:
!bash ../xvfb start
os.environ['DISPLAY'] = ':1'
import gym
import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
%matplotlib inline
env = gym.make("CartPole-v0"... | [2017-03-14 19:35:59,320] Making new env: CartPole-v0
| Unlicense | week06_policy_based/reinforce_lasagne.ipynb | kianya/Practical_RL |
Building the network for REINFORCE For REINFORCE algorithm, we'll need a model that predicts action probabilities given states. | import theano
import theano.tensor as T
# create input variables. We'll support multiple states at once
states = T.matrix("states[batch,units]")
actions = T.ivector("action_ids[batch]")
cumulative_rewards = T.vector("G[batch] = r + gamma*r' + gamma^2*r'' + ...")
import lasagne
from lasagne.layers import *
# input la... | _____no_output_____ | Unlicense | week06_policy_based/reinforce_lasagne.ipynb | kianya/Practical_RL |
Predict function | # get probabilities of actions
predicted_probas = get_output(l_action_probas)
# predict action probability given state
# if you use float32, set allow_input_downcast=True
predict_proba = <compile a function that takes states and returns predicted_probas > | _____no_output_____ | Unlicense | week06_policy_based/reinforce_lasagne.ipynb | kianya/Practical_RL |
Loss function and updatesWe now need to define objective and update over policy gradient.Our objective function is$$ J \approx { 1 \over N } \sum _{s_i,a_i} \pi_\theta (a_i | s_i) \cdot G(s_i,a_i) $$Following the REINFORCE algorithm, we can define our objective as follows: $$ \hat J \approx { 1 \over N } \sum _{s_i... | # select probabilities for chosen actions, pi(a_i|s_i)
predicted_probas_for_actions = predicted_probas[T.arange(
actions.shape[0]), actions]
# REINFORCE objective function
J = # <policy objective as in the last formula. Please use mean, not sum.>
# all network weights
all_weights = <get all "thetas" aka network we... | _____no_output_____ | Unlicense | week06_policy_based/reinforce_lasagne.ipynb | kianya/Practical_RL |
Computing cumulative rewards |
def get_cumulative_rewards(rewards, # rewards at each step
gamma=0.99 # discount for reward
):
"""
take a list of immediate rewards r(s,a) for the whole session
compute cumulative returns (a.k.a. G(s,a) in Sutton '16)
G_t = r_t + gamma*r_{t+1} +... | _____no_output_____ | Unlicense | week06_policy_based/reinforce_lasagne.ipynb | kianya/Practical_RL |
Playing the game | def generate_session(t_max=1000):
"""play env with REINFORCE agent and train at the session end"""
# arrays to record session
states, actions, rewards = [], [], []
s = env.reset()
for t in range(t_max):
# action probabilities array aka pi(a|s)
action_probas = predict_proba([s])[0... | mean reward:20.900
mean reward:35.860
mean reward:50.820
mean reward:88.550
mean reward:132.080
mean reward:165.890
mean reward:193.790
mean reward:166.510
mean reward:120.910
mean reward:98.450
mean reward:236.340
mean reward:280.410
mean reward:317.610
You Win!
| Unlicense | week06_policy_based/reinforce_lasagne.ipynb | kianya/Practical_RL |
Video | # record sessions
import gym.wrappers
env = gym.wrappers.Monitor(gym.make("CartPole-v0"),
directory="videos", force=True)
sessions = [generate_session() for _ in range(100)]
env.close()
# show video
from IPython.display import HTML
import os
video_names = list(
filter(lambda s: s.endswit... | _____no_output_____ | Unlicense | week06_policy_based/reinforce_lasagne.ipynb | kianya/Practical_RL |
Reflect Tables into SQLALchemy ORM | # Python SQL toolkit and Object Relational Mapper
import sqlalchemy
from sqlalchemy.ext.automap import automap_base
from sqlalchemy.orm import Session
from sqlalchemy import create_engine, func, inspect
# create engine to hawaii.sqlite
engine = create_engine("sqlite:///./Resources/hawaii.sqlite")
# reflect an existing ... | _____no_output_____ | MIT | .ipynb_checkpoints/temp_analysis_bonus_2_starter - Copy-checkpoint.ipynb | pratixashah/sqlalchemy |
Bonus Challenge Assignment: Temperature Analysis II | # This function called `calc_temps` will accept start date and end date in the format '%Y-%m-%d'
# and return the minimum, maximum, and average temperatures for that range of dates
def calc_temps(start_date, end_date):
"""TMIN, TAVG, and TMAX for a list of dates.
Args:
start_date (string): A date ... | _____no_output_____ | MIT | .ipynb_checkpoints/temp_analysis_bonus_2_starter - Copy-checkpoint.ipynb | pratixashah/sqlalchemy |
Daily Rainfall Average | # Calculate the total amount of rainfall per weather station for your trip dates using the previous year's
# matching dates.
# Sort this in descending order by precipitation amount and list the station, name, latitude, longitude, and elevation
# Use this function to calculate the daily normals
# (i.e. the averages f... | _____no_output_____ | MIT | .ipynb_checkpoints/temp_analysis_bonus_2_starter - Copy-checkpoint.ipynb | pratixashah/sqlalchemy |
📝 Exercise M3.02The goal is to find the best set of hyperparameters which maximize thegeneralization performance on a training set.Here again with limit the size of the training set to make computationrun faster. Feel free to increase the `train_size` value if your computeris powerful enough. |
import numpy as np
import pandas as pd
adult_census = pd.read_csv("../datasets/adult-census.csv")
target_name = "class"
target = adult_census[target_name]
data = adult_census.drop(columns=[target_name, "education-num"])
from sklearn.model_selection import train_test_split
data_train, data_test, target_train, target... | _____no_output_____ | MIT | sklearn/notes/parameter_tuning_ex_03.ipynb | shamik-biswas-rft/CodeSnippets |
In this exercise, we will progressively define the classification pipelineand later tune its hyperparameters.Our pipeline should:* preprocess the categorical columns using a `OneHotEncoder` and use a `StandardScaler` to normalize the numerical data.* use a `LogisticRegression` as a predictive model.Start by defining t... | from sklearn.compose import make_column_selector as selector
# Write your code here.
categorical_selector = selector(dtype_include=object)
numerical_selector = selector(dtype_exclude=object)
categorical_columns = categorical_selector(data)
numerical_columns = numerical_selector(data)
from sklearn.preprocessing import... | _____no_output_____ | MIT | sklearn/notes/parameter_tuning_ex_03.ipynb | shamik-biswas-rft/CodeSnippets |
Subsequently, create a `ColumnTransformer` to redirect the specific columnsa preprocessing pipeline. | from sklearn.compose import ColumnTransformer
# Write your code here.
preprocessor = ColumnTransformer(
[
('cat_process', cat_processor, categorical_columns),
('num_process', num_processor, numerical_columns)
]) | _____no_output_____ | MIT | sklearn/notes/parameter_tuning_ex_03.ipynb | shamik-biswas-rft/CodeSnippets |
Assemble the final pipeline by combining the above preprocessorwith a logistic regression classifier. Force the maximum number ofiterations to `10_000` to ensure that the model will converge. | from sklearn.pipeline import make_pipeline
from sklearn.linear_model import LogisticRegression
# Write your code here.
model = make_pipeline(preprocessor, LogisticRegression(max_iter=11_000)) | _____no_output_____ | MIT | sklearn/notes/parameter_tuning_ex_03.ipynb | shamik-biswas-rft/CodeSnippets |
Use `RandomizedSearchCV` with `n_iter=20` to find the best set ofhyperparameters by tuning the following parameters of the `model`:- the parameter `C` of the `LogisticRegression` with values ranging from 0.001 to 10. You can use a log-uniform distribution (i.e. `scipy.stats.loguniform`);- the parameter `with_mean` of... | model.get_params()
from sklearn.model_selection import RandomizedSearchCV
from scipy.stats import loguniform
# Write your code here.
params_dict = {
'columntransformer__num_process__with_mean': [True, False],
'columntransformer__num_process__with_std': [True, False],
'logisticregression__C': loguniform(1e-... | Fitting 5 folds for each of 20 candidates, totalling 100 fits
| MIT | sklearn/notes/parameter_tuning_ex_03.ipynb | shamik-biswas-rft/CodeSnippets |
Stage 3: What do I need to install?Maybe your experience looks like the typical python dependency management (https://xkcd.com/1987/):Furthermore, data science packages can have all sorts of additional non-Python dependencies which makes things even more confusing, and we end up spending more time sorting out our depe... | from sklearn.preprocessing import StandardScaler
import matplotlib.pyplot as plt
import pandas as pd
%matplotlib inline | _____no_output_____ | MIT | notebooks/03-Environment-Challenge.ipynb | acwooding/easydata-tutorial |
Updating your conda and pip environmentsThe `make` commands, `make create_environment` and `make update_environment` are wrappers that allow you to easily manage your conda and pip environments using a file called `environment.yml` file, which lists the packages you want in your python environment.(If you ever forget ... | %conda info | _____no_output_____ | MIT | notebooks/03-Environment-Challenge.ipynb | acwooding/easydata-tutorial |
Another useful cell to include is the following.If you want your environment changes to be immediately available in your running notebooks, make sure to run a notebook cell containing: | %load_ext autoreload
%autoreload 2 | _____no_output_____ | MIT | notebooks/03-Environment-Challenge.ipynb | acwooding/easydata-tutorial |
If you did your task correctly, the following import will succeed. | import seaborn as sns | _____no_output_____ | MIT | notebooks/03-Environment-Challenge.ipynb | acwooding/easydata-tutorial |
- First make sure that our notebook can display a simple widget. Some ppl might have trouble with this. | from ipywidgets import interact
def f(x):
return x
interact(f, x=10) | _____no_output_____ | MIT | demo.ipynb | w-hc/coco_vis |
- If you see a slider that can be dragged, it's good. - Resolve any widget related notebook setup problems before proceeding | vis = Visualizer(
coco_eval, matching_threshold=0.5, area='all', img_directory=img_directory
) | _____no_output_____ | MIT | demo.ipynb | w-hc/coco_vis |
- the widget will disappear after each session. Have to re-run | vis.category_PR_curve()
vis.image_vis() | _____no_output_____ | MIT | demo.ipynb | w-hc/coco_vis |
Load Data | key = data.get_output_keys()[2]
key
output = data.load_output(key)
qf = torch.Tensor(output["qf"])
gf = torch.Tensor(output["gf"])
q_pids = np.array(output["q_pids"])
g_pids = np.array(output["g_pids"])
q_camids = np.array(output["q_camids"])
g_camids = np.array(output["g_camids"])
distmat = compute_distmat(qf, gf) | _____no_output_____ | MIT | notebooks/Post Rank.ipynb | itsnamgyu/reid-metric |
Baseline Results | result = evaluate(distmat, q_pids, g_pids, q_camids, g_camids)
result | _____no_output_____ | MIT | notebooks/Post Rank.ipynb | itsnamgyu/reid-metric |
Re-ranked Results | all_distmat = compute_inner_distmat(torch.cat((qf, gf)))
re_distmat = rerank_distmat(all_distmat, qf.shape[0])
re_result = evaluate(re_distmat, q_pids, g_pids, q_camids, g_camids)
re_result | _____no_output_____ | MIT | notebooks/Post Rank.ipynb | itsnamgyu/reid-metric |
One-Shot Evaluation Using Modules Rocchio | ntot = torch.as_tensor
rocchio.run(qf, gf, ntot(q_pids), ntot(g_pids), ntot(q_camids), ntot(g_camids), t=3) | _____no_output_____ | MIT | notebooks/Post Rank.ipynb | itsnamgyu/reid-metric |
Neighborhood Expansion (Min) | ntot = torch.from_numpy
ne.run(qf, gf, ntot(q_pids), ntot(g_pids), ntot(q_camids), ntot(g_camids), t=3, method="min") | _____no_output_____ | MIT | notebooks/Post Rank.ipynb | itsnamgyu/reid-metric |
Neighborhood Expansion (Mean) | ntot = torch.from_numpy
ne.run(qf, gf, ntot(q_pids), ntot(g_pids), ntot(q_camids), ntot(g_camids), t=3, method="mean") | _____no_output_____ | MIT | notebooks/Post Rank.ipynb | itsnamgyu/reid-metric |
Development Feedback Models | q_pids = torch.tensor(output["q_pids"])
g_pids = torch.tensor(output["g_pids"])
q = len(q_pids)
g = len(g_pids)
m = qf.shape[1]
q_camids = np.array(output["q_camids"])
g_camids = np.array(output["g_camids"]) | _____no_output_____ | MIT | notebooks/Post Rank.ipynb | itsnamgyu/reid-metric |
Naive Feedback | if input("reset? ") == "y":
positive_indices = torch.zeros((q, g), dtype=bool)
negative_indices = torch.zeros((q, g), dtype=bool)
for i in tqdm(range(5)):
qf_adjusted = qf # no adjust, naive re-rank
distmat = compute_distmat(qf_adjusted, gf)
distmat[positive_indices] = float("inf")
distmat[neg... | _____no_output_____ | MIT | notebooks/Post Rank.ipynb | itsnamgyu/reid-metric |
Rocchio | if input("reset? ") == "y":
positive_indices = torch.zeros((q, g), dtype=bool)
negative_indices = torch.zeros((q, g), dtype=bool)
alpha = 1
beta = 0.65
gamma = 0.35
qf_adjusted = qf
for i in tqdm(range(5)):
distmat = compute_distmat(qf_adjusted, gf)
distmat[positive_indices] = float("inf")
distmat[... | _____no_output_____ | MIT | notebooks/Post Rank.ipynb | itsnamgyu/reid-metric |
Function Tests (Rocchio) | def adjust_qf(qf, gf, positive_indices, negative_indices, alpha=1, beta=0.65, gamma=0.35):
assert(qf.shape[1] == gf.shape[1])
mean_positive_gf = positive_indices.float().mm(gf) / positive_indices.float().sum(dim=1, keepdim=True)
mean_negative_gf = negative_indices.float().mm(gf) / negative_indices.float().s... | _____no_output_____ | MIT | notebooks/Post Rank.ipynb | itsnamgyu/reid-metric |
Module Test (Naive) | positive_indices = None
negative_indices = None
distmat = None
for i in tqdm(range(3)):
distmat, positive_indices, negative_indices = feedback.naive_round(
qf, gf, q_pids, g_pids, positive_indices, negative_indices, previous_distmat=distmat)
naive_result = evaluate(distmat, q_pids, g_pids, q_camids, g_camid... | _____no_output_____ | MIT | notebooks/Post Rank.ipynb | itsnamgyu/reid-metric |
Module Test (Rocchio) | positive_indices = None
negative_indices = None
distmat = None
for i in tqdm(range(3)):
distmat, positive_indices, negative_indices = rocchio.rocchio_round(
qf, gf, q_pids, g_pids, positive_indices, negative_indices, previous_distmat=distmat)
rocchio_result = evaluate(distmat, q_pids, g_pids, q_camids, g_ca... | _____no_output_____ | MIT | notebooks/Post Rank.ipynb | itsnamgyu/reid-metric |
Single Feedback Rocchio (Old)Initial implementation test | g_pids = torch.Tensor(output["g_pids"])
q_pids = torch.Tensor(output["q_pids"])
match = g_pids[min_indices] == q_pids
selected_gf = gf[min_indices]
selected_gf.shape
weights = match.float() * (beta + gamma) - gamma
weighted_feedback = selected_gf * weights.reshape(-1, 1)
weighted_feedback
inverse_weights = 1 - weights
... | _____no_output_____ | MIT | notebooks/Post Rank.ipynb | itsnamgyu/reid-metric |
View source on GitHub Notebook Viewer Run in binder Run in Google Colab Install Earth Engine APIInstall the [Earth Engine Python API](https://developers.google.com/earth-engine/python_install) and [geehydro](https://github.com/giswqs/geehydro). The **geehydro** Python package builds on the [folium](http... | # %%capture
# !pip install earthengine-api
# !pip install geehydro | _____no_output_____ | MIT | Image/texture.ipynb | guy1ziv2/earthengine-py-notebooks |
Import libraries | import ee
import folium
import geehydro | _____no_output_____ | MIT | Image/texture.ipynb | guy1ziv2/earthengine-py-notebooks |
Authenticate and initialize Earth Engine API. You only need to authenticate the Earth Engine API once. Uncomment the line `ee.Authenticate()` if you are running this notebook for this first time or if you are getting an authentication error. | # ee.Authenticate()
ee.Initialize() | _____no_output_____ | MIT | Image/texture.ipynb | guy1ziv2/earthengine-py-notebooks |
Create an interactive map This step creates an interactive map using [folium](https://github.com/python-visualization/folium). The default basemap is the OpenStreetMap. Additional basemaps can be added using the `Map.setOptions()` function. The optional basemaps can be `ROADMAP`, `SATELLITE`, `HYBRID`, `TERRAIN`, or `... | Map = folium.Map(location=[40, -100], zoom_start=4)
Map.setOptions('HYBRID') | _____no_output_____ | MIT | Image/texture.ipynb | guy1ziv2/earthengine-py-notebooks |
Add Earth Engine Python script | import math
# Load a high-resolution NAIP image.
image = ee.Image('USDA/NAIP/DOQQ/m_3712213_sw_10_1_20140613')
# Zoom to San Francisco, display.
Map.setCenter(-122.466123, 37.769833, 17)
Map.addLayer(image, {'max': 255}, 'image')
# Get the NIR band.
nir = image.select('N')
# Define a neighborhood with a kernel.
squ... | _____no_output_____ | MIT | Image/texture.ipynb | guy1ziv2/earthengine-py-notebooks |
Display Earth Engine data layers | Map.setControlVisibility(layerControl=True, fullscreenControl=True, latLngPopup=True)
Map | _____no_output_____ | MIT | Image/texture.ipynb | guy1ziv2/earthengine-py-notebooks |
TFMA Notebook exampleThis notebook describes how to export your model for TFMA and demonstrates the analysis tooling it offers.Note: Please make sure to follow the instructions in [README.md](https://github.com/tensorflow/tfx/blob/master/tfx/examples/chicago_taxi/README.md) when running this notebook SetupImport neces... | import apache_beam as beam
import os
import preprocess
import shutil
import tensorflow as tf
import tensorflow_data_validation as tfdv
import tensorflow_model_analysis as tfma
from google.protobuf import text_format
from tensorflow.python.lib.io import file_io
from tensorflow_transform.beam.tft_beam_io import transf... | _____no_output_____ | Apache-2.0 | tfx/examples/chicago_taxi/chicago_taxi_tfma_local_playground.ipynb | yongsheng268/tfx |
Helper functions and some constants for running the notebook locally. | BASE_DIR = os.getcwd()
DATA_DIR = os.path.join(BASE_DIR, 'data')
OUTPUT_DIR = os.path.join(BASE_DIR, 'chicago_taxi_output')
# Base dir containing train and eval data
TRAIN_DATA_DIR = os.path.join(DATA_DIR, 'train')
EVAL_DATA_DIR = os.path.join(DATA_DIR, 'eval')
# Base dir where TFT writes training data
TFT_TRAIN_OU... | _____no_output_____ | Apache-2.0 | tfx/examples/chicago_taxi/chicago_taxi_tfma_local_playground.ipynb | yongsheng268/tfx |
Clean up output directories. | shutil.rmtree(TFT_TRAIN_OUTPUT_BASE_DIR, ignore_errors=True)
shutil.rmtree(TFT_EVAL_OUTPUT_BASE_DIR, ignore_errors=True)
shutil.rmtree(TF_OUTPUT_BASE_DIR, ignore_errors=True)
shutil.rmtree(get_schema_file(), ignore_errors=True) | _____no_output_____ | Apache-2.0 | tfx/examples/chicago_taxi/chicago_taxi_tfma_local_playground.ipynb | yongsheng268/tfx |
Compute and visualize descriptive data statistics | # Compute stats over training data.
train_stats = tfdv.generate_statistics_from_csv(data_location=os.path.join(TRAIN_DATA_DIR, 'data.csv'))
# Visualize training data stats.
tfdv.visualize_statistics(train_stats) | _____no_output_____ | Apache-2.0 | tfx/examples/chicago_taxi/chicago_taxi_tfma_local_playground.ipynb | yongsheng268/tfx |
Infer a schema | # Infer a schema from the training data stats.
schema = tfdv.infer_schema(statistics=train_stats, infer_feature_shape=False)
tfdv.display_schema(schema=schema) | _____no_output_____ | Apache-2.0 | tfx/examples/chicago_taxi/chicago_taxi_tfma_local_playground.ipynb | yongsheng268/tfx |
Check evaluation data for errors | # Compute stats over eval data.
eval_stats = tfdv.generate_statistics_from_csv(data_location=os.path.join(EVAL_DATA_DIR, 'data.csv'))
# Compare stats of eval data with training data.
tfdv.visualize_statistics(lhs_statistics=eval_stats, rhs_statistics=train_stats,
lhs_name='EVAL_DATASET', rhs_n... | _____no_output_____ | Apache-2.0 | tfx/examples/chicago_taxi/chicago_taxi_tfma_local_playground.ipynb | yongsheng268/tfx |
Freeze the schemaNow that the schema has been reviewed and curated, we will store it in a file to reflect its "frozen" state. | file_io.recursive_create_dir(OUTPUT_DIR)
file_io.write_string_to_file(get_schema_file(), text_format.MessageToString(schema)) | _____no_output_____ | Apache-2.0 | tfx/examples/chicago_taxi/chicago_taxi_tfma_local_playground.ipynb | yongsheng268/tfx |
Preprocess Inputstransform_data is defined in preprocess.py and uses the tensorflow_transform library to perform preprocessing. The same code is used for both local preprocessing in this notebook and preprocessing in the Cloud (via Dataflow). | # Transform eval data
preprocess.transform_data(input_handle=os.path.join(EVAL_DATA_DIR, 'data.csv'),
outfile_prefix=TFT_EVAL_FILE_PREFIX,
working_dir=get_tft_eval_output_dir(0),
schema_file=get_schema_file(),
pipel... | _____no_output_____ | Apache-2.0 | tfx/examples/chicago_taxi/chicago_taxi_tfma_local_playground.ipynb | yongsheng268/tfx |
Compute statistics over transformed data | # Compute stats over transformed training data.
TRANSFORMED_TRAIN_DATA = os.path.join(get_tft_train_output_dir(0), TFT_TRAIN_FILE_PREFIX + "*")
transformed_train_stats = tfdv.generate_statistics_from_tfrecord(data_location=TRANSFORMED_TRAIN_DATA)
# Visualize transformed training data stats and compare to raw training ... | _____no_output_____ | Apache-2.0 | tfx/examples/chicago_taxi/chicago_taxi_tfma_local_playground.ipynb | yongsheng268/tfx |
Prepare the ModelTo use TFMA, export the model into an **EvalSavedModel** by calling ``tfma.export.export_eval_savedmodel``.``tfma.export.export_eval_savedmodel`` is analogous to ``estimator.export_savedmodel`` but exports the evaluation graph as opposed to the training or inference graph. Notice that one of the input... | def run_experiment(hparams):
"""Run the training and evaluate using the high level API"""
# Train and evaluate the model as usual.
estimator = task.train_and_maybe_evaluate(hparams)
# Export TFMA's sepcial EvalSavedModel
eval_model_dir = os.path.join(hparams.output_dir, EVAL_MODEL_DIR)
receive... | _____no_output_____ | Apache-2.0 | tfx/examples/chicago_taxi/chicago_taxi_tfma_local_playground.ipynb | yongsheng268/tfx |
Train and export the model for TFMA | def run_local_experiment(tft_run_id, tf_run_id, num_layers, first_layer_size, scale_factor):
"""Helper method to train and export the model for TFMA
The caller specifies the input and output directory by providing run ids. The optional parameters
allows the user to change the modelfor time series view.... | _____no_output_____ | Apache-2.0 | tfx/examples/chicago_taxi/chicago_taxi_tfma_local_playground.ipynb | yongsheng268/tfx |
Run TFMA to compute metricsFor local analysis, TFMA offers a helper method ``tfma.run_model_analysis`` | help(tfma.run_model_analysis) | _____no_output_____ | Apache-2.0 | tfx/examples/chicago_taxi/chicago_taxi_tfma_local_playground.ipynb | yongsheng268/tfx |
You can also write your own custom pipeline if you want to perform extra transformations on the data before evaluation. | def run_tfma(slice_spec, tf_run_id, tfma_run_id, input_csv, schema_file, add_metrics_callbacks=None):
"""A simple wrapper function that runs tfma locally.
A function that does extra transformations on the data and then run model analysis.
Args:
slice_spec: The slicing spec for how to slice... | _____no_output_____ | Apache-2.0 | tfx/examples/chicago_taxi/chicago_taxi_tfma_local_playground.ipynb | yongsheng268/tfx |
You can also compute metrics on slices of your data in TFMA. Slices can be specified using ``tfma.slicer.SingleSliceSpec``.Below are examples of how slices can be specified. | # An empty slice spec means the overall slice, that is, the whole dataset.
OVERALL_SLICE_SPEC = tfma.slicer.SingleSliceSpec()
# Data can be sliced along a feature column
# In this case, data is sliced along feature column trip_start_hour.
FEATURE_COLUMN_SLICE_SPEC = tfma.slicer.SingleSliceSpec(columns=['trip_start_hou... | _____no_output_____ | Apache-2.0 | tfx/examples/chicago_taxi/chicago_taxi_tfma_local_playground.ipynb | yongsheng268/tfx |
Let's run TFMA! | tf.logging.set_verbosity(tf.logging.INFO)
tfma_result_1 = run_tfma(input_csv=os.path.join(EVAL_DATA_DIR, 'data.csv'),
tf_run_id=0,
tfma_run_id=1,
slice_spec=ALL_SPECS,
schema_file=get_schema_file())
print('Done')
| _____no_output_____ | Apache-2.0 | tfx/examples/chicago_taxi/chicago_taxi_tfma_local_playground.ipynb | yongsheng268/tfx |
Visualization: Slicing MetricsTo see the slices, either use the name of the column (by setting slicing_column) or provide a tfma.slicer.SingleSliceSpec (by setting slicing_spec). If neither is provided, the overall will be displayed.The default visualization is **slice overview** when the number of slices is small. It... | # Show data sliced along feature column trip_start_hour.
tfma.view.render_slicing_metrics(
tfma_result_1, slicing_column='trip_start_hour')
# Show metrics sliced by COLUMN_CROSS_VALUE_SPEC above.
tfma.view.render_slicing_metrics(tfma_result_1, slicing_spec=COLUMN_CROSS_VALUE_SPEC)
# Show overall metrics.
tfma.view.... | _____no_output_____ | Apache-2.0 | tfx/examples/chicago_taxi/chicago_taxi_tfma_local_playground.ipynb | yongsheng268/tfx |
Visualization: PlotsTFMA offers a number of built-in plots. To see them, add them to ``add_metrics_callbacks`` | tf.logging.set_verbosity(tf.logging.INFO)
tfma_vis = run_tfma(input_csv=os.path.join(EVAL_DATA_DIR, 'data.csv'),
tf_run_id=0,
tfma_run_id='vis',
slice_spec=ALL_SPECS,
schema_file=get_schema_file(),
add_metrics_callback... | _____no_output_____ | Apache-2.0 | tfx/examples/chicago_taxi/chicago_taxi_tfma_local_playground.ipynb | yongsheng268/tfx |
Plots must be visualized for an individual slice. To specify a slice, use ``tfma.slicer.SingleSliceSpec``.In the example below, we are using ``tfma.slicer.SingleSliceSpec(features=[('trip_start_hour', 1)])`` to specify the slice where trip_start_hour is 1.Plots are interactive:- Drag to pan- Scroll to zoom- Right click... | tfma.view.render_plot(tfma_vis, tfma.slicer.SingleSliceSpec(features=[('trip_start_hour', 1)])) | _____no_output_____ | Apache-2.0 | tfx/examples/chicago_taxi/chicago_taxi_tfma_local_playground.ipynb | yongsheng268/tfx |
Custom metricsIn addition to plots, it is also possible to compute additional metrics not present at export time or custom metrics metrics using ``add_metrics_callbacks``.All metrics in ``tf.metrics`` are supported in the callback and can be used to compose other metrics:https://www.tensorflow.org/api_docs/python/tf/m... | # Defines a callback that adds FNR to the result.
def add_fnr_for_threshold(threshold):
def _add_fnr_callback(features_dict, predictions_dict, labels_dict):
metric_ops = {}
prediction_tensor = tf.cast(
predictions_dict.get(tf.contrib.learn.PredictionKey.LOGISTIC), tf.float64)
fn_... | _____no_output_____ | Apache-2.0 | tfx/examples/chicago_taxi/chicago_taxi_tfma_local_playground.ipynb | yongsheng268/tfx |
Visualization: Time SeriesIt is important to track how your model is doing over time. TFMA offers two modes to show your model performs over time.**Multiple model analysis** shows how model perfoms from one version to another. This is useful early on to see how the addition of new features, change in modeling techniqu... | help(tfma.multiple_model_analysis) | _____no_output_____ | Apache-2.0 | tfx/examples/chicago_taxi/chicago_taxi_tfma_local_playground.ipynb | yongsheng268/tfx |
**Multiple data analysis** shows how a model perfoms under different evaluation data set. This is useful to ensure that model performance does not degrade over time. TFMA offer a conveneient method. | help(tfma.multiple_data_analysis) | _____no_output_____ | Apache-2.0 | tfx/examples/chicago_taxi/chicago_taxi_tfma_local_playground.ipynb | yongsheng268/tfx |
It is also possible to compose a time series manually. | # Create different models.
# Run some experiments with different hidden layer configurations.
run_local_experiment(tft_run_id=0,
tf_run_id=1,
num_layers=3,
first_layer_size=200,
scale_factor=0.7)
run_local_experiment(tft_run_id=0,
... | _____no_output_____ | Apache-2.0 | tfx/examples/chicago_taxi/chicago_taxi_tfma_local_playground.ipynb | yongsheng268/tfx |
Like plots, time series view must visualized for a slice too.In the example below, we are showing the overall slice.Select a metric to see its time series graph. Hover over each data point to get more details. | eval_results = tfma.make_eval_results([tfma_result_1, tfma_result_2, tfma_result_3],
tfma.constants.MODEL_CENTRIC_MODE)
tfma.view.render_time_series(eval_results, OVERALL_SLICE_SPEC)
| _____no_output_____ | Apache-2.0 | tfx/examples/chicago_taxi/chicago_taxi_tfma_local_playground.ipynb | yongsheng268/tfx |
Serialized results can also be used to construct a time series. Thus, there is no need to re-run TFMA for models already evaluated for a long running pipeline. | # Visualize the results in a Time Series. In this case, we are showing the slice specified.
eval_results_from_disk = tfma.load_eval_results([get_tfma_output_dir(1),
get_tfma_output_dir(2),
get_tfma_output_dir(3)],
... | _____no_output_____ | Apache-2.0 | tfx/examples/chicago_taxi/chicago_taxi_tfma_local_playground.ipynb | yongsheng268/tfx |
Conservative remapping | import xgcm
import xarray as xr
import numpy as np
import xbasin | _____no_output_____ | MIT | docs/remapping.ipynb | BenniSchmiedel/ECO |
We open the example data and create 2 grids: 1 for the dataset we have and 1 for the remapped one.Here '_fr' means *from* and '_to' *to* (i.e. remapped data). | ds = xr.open_dataset('data/nemo_output_ex.nc')
from xnemogcm import open_nemo_and_domain_cfg
ds = open_nemo_and_domain_cfg(datadir='/home/romain/Documents/Education/PhD/Courses/2019-OC6310/Project/Experiments/EXP_eos00/Rawdata')
metrics_fr = {
('X',): ['e1t', 'e1u', 'e1v', 'e1f'],
('Y',): ['e2t', 'e2u', 'e2v'... | <xarray.Dataset>
Dimensions: (axis_nbounds: 2, t: 5, x_c: 21, x_f: 21, y_c: 40, y_f: 40, z_c: 36, z_f: 36)
Coordinates:
* z_f (z_f) float64 -0.5 0.5 1.5 2.5 3.5 ... 31.5 32.5 33.5 34.5
* t (t) object 7756-01-01 00:00:00 ... 7796-01-01 00:00:00
* x_c (x_c) int64 0 1 2 3 4 ... | MIT | docs/remapping.ipynb | BenniSchmiedel/ECO |
Remap a T point | %timeit xbasin.remap_vertical(ds.thetao, grid_fr, grid_to, axis='Z')
theta_to = xbasin.remap_vertical(ds.thetao, grid_fr, grid_to, axis='Z')
print(theta_to.coords) | Coordinates:
* z_c (z_c) int64 0 1 2 3 4 5 6 7 8 9 ... 26 27 28 29 30 31 32 33 34 35
* t (t) object 7756-01-01 00:00:00 ... 7796-01-01 00:00:00
* x_c (x_c) int64 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
* y_c (y_c) int64 0 1 2 3 4 5 6 7 8 9 ... 30 31 32 33 34 35 36 37 38 39
| MIT | docs/remapping.ipynb | BenniSchmiedel/ECO |
The total heat content is conserved: | hc_fr = grid_fr.integrate(ds.thetao, axis='Z')
hc_to = grid_to.integrate(theta_to, axis='Z')
(hc_fr == hc_to).all() | _____no_output_____ | MIT | docs/remapping.ipynb | BenniSchmiedel/ECO |
Remap a W point | w_to = xbasin.remap_vertical(ds.woce*0+1, grid_fr, grid_to, axis='Z')
grid_to.integrate(w_to, axis='Z')[-1].plot()
grid_fr.integrate((ds.woce*0+1), axis='Z')[-1].plot() | _____no_output_____ | MIT | docs/remapping.ipynb | BenniSchmiedel/ECO |
Time comparisonThe heart function of the remapping is computed from python to C++ with pythran, which improves the speed. However if pythran is not installed, the original python function is called instead.As a user, you should not use the 2 following functions, they are only shown here for the time comparison. | fake_dataset = [
np.ascontiguousarray(ds.gdept_0.values.reshape(ds.gdept_0.values.shape+(1,))),
np.ascontiguousarray(ds.gdepw_0.values.reshape(ds.gdepw_0.values.shape+(1,))),
np.ascontiguousarray(ds.thetao.transpose('z_c', 'y_c', 'x_c', 't').values.flatten().reshape(ds.thetao.transpose('z_c', 'y_c', 'x_c', ... | _____no_output_____ | MIT | docs/remapping.ipynb | BenniSchmiedel/ECO |
Pure Python | %timeit _interpolation_pure_python(*fake_dataset) | 120 ms ± 1.1 ms per loop (mean ± std. dev. of 7 runs, 10 loops each)
| MIT | docs/remapping.ipynb | BenniSchmiedel/ECO |
Pythran | %timeit _interpolation_pythran(*fake_dataset) | 2.66 ms ± 16.2 µs per loop (mean ± std. dev. of 7 runs, 100 loops each)
| MIT | docs/remapping.ipynb | BenniSchmiedel/ECO |
DeepDreaming with TensorFlow >[Loading and displaying the model graph](loading)>[Naive feature visualization](naive)>[Multiscale image generation](multiscale)>[Laplacian Pyramid Gradient Normalization](laplacian)>[Playing with feature visualzations](playing)>[DeepDream](deepdream) This notebook demonstrates a number o... | # boilerplate code
from __future__ import print_function
import os
from io import BytesIO
import numpy as np
from functools import partial
import PIL.Image
from IPython.display import clear_output, Image, display, HTML
import tensorflow as tf | _____no_output_____ | Apache-2.0 | tensorflow/examples/tutorials/deepdream/deepdream.ipynb | espuer/tensorflow |
Loading and displaying the model graphThe pretrained network can be downloaded [here](https://storage.googleapis.com/download.tensorflow.org/models/inception5h.zip). Unpack the `tensorflow_inception_graph.pb` file from the archive and set its path to `model_fn` variable. Alternatively you can uncomment and run the fol... | #!wget https://storage.googleapis.com/download.tensorflow.org/models/inception5h.zip && unzip inception5h.zip
model_fn = 'tensorflow_inception_graph.pb'
# creating TensorFlow session and loading the model
graph = tf.Graph()
sess = tf.InteractiveSession(graph=graph)
with tf.gfile.FastGFile(model_fn, 'rb') as f:
gra... | _____no_output_____ | Apache-2.0 | tensorflow/examples/tutorials/deepdream/deepdream.ipynb | espuer/tensorflow |
To take a glimpse into the kinds of patterns that the network learned to recognize, we will try to generate images that maximize the sum of activations of particular channel of a particular convolutional layer of the neural network. The network we explore contains many convolutional layers, each of which outputs tens t... | layers = [op.name for op in graph.get_operations() if op.type=='Conv2D' and 'import/' in op.name]
feature_nums = [int(graph.get_tensor_by_name(name+':0').get_shape()[-1]) for name in layers]
print('Number of layers', len(layers))
print('Total number of feature channels:', sum(feature_nums))
# Helper functions for TF... | Number of layers 59
Total number of feature channels: 7548
| Apache-2.0 | tensorflow/examples/tutorials/deepdream/deepdream.ipynb | espuer/tensorflow |
Naive feature visualization Let's start with a naive way of visualizing these. Image-space gradient ascent! | # Picking some internal layer. Note that we use outputs before applying the ReLU nonlinearity
# to have non-zero gradients for features with negative initial activations.
layer = 'mixed4d_3x3_bottleneck_pre_relu'
channel = 139 # picking some feature channel to visualize
# start with a gray image with a little noise
im... | _____no_output_____ | Apache-2.0 | tensorflow/examples/tutorials/deepdream/deepdream.ipynb | espuer/tensorflow |
Multiscale image generationLooks like the network wants to show us something interesting! Let's help it. We are going to apply gradient ascent on multiple scales. Details formed on smaller scale will be upscaled and augmented with additional details on the next scale.With multiscale image generation it may be tempting... | def tffunc(*argtypes):
'''Helper that transforms TF-graph generating function into a regular one.
See "resize" function below.
'''
placeholders = list(map(tf.placeholder, argtypes))
def wrap(f):
out = f(*placeholders)
def wrapper(*args, **kw):
return out.eval(dict(zip(pla... | _____no_output_____ | Apache-2.0 | tensorflow/examples/tutorials/deepdream/deepdream.ipynb | espuer/tensorflow |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.