Friday, 11 October 2019
Thursday, 10 October 2019
Tuesday, 8 October 2019
react hooks createContext
useContext enables sharing state between multiple components without explicitly passing a prop through every level of the tree.
reference:
https://daveceddia.com/usecontext-hook/
https://www.codementor.io/sambhavgore/an-example-use-context-and-hooks-to-share-state-between-different-components-sgop6lnrd
part initially named screw,
part name and function to change name are stored in context created at top level.
they are accessible to lower levels
when part name is changed at any level, part name on other level auto updated as well
no passing props from top down required
//Top
import React, { createContext, useState } from "react";
import SubAssembly from './Sub'
// Create a Context
export const partContext = createContext();
// It returns an object with 2 values:
// { Provider, Consumer }
function TopAssembly() {
// Use the Provider to make a value available to all
// children and grandchildren
const [partName, setPartName] = useState('screw ')
return (
<partContext.Provider value={{ partName, setPartName }}>
<div>
Top Assembly <button onClick={() => setPartName('screw A ')}>update part on top level</button>
<ul>
<li>sub Assembly x</li>
<li>sub Assembly y</li>
<li><SubAssembly /></li>
</ul>
</div>
</partContext.Provider>
);
}
export default TopAssembly
------------------------------------
//Sub
import React, {useContext } from "react";
import {partContext} from './Top'
import BaseAssembly from './Base'
function SubAssembly() {
const { setPartName } = useContext(partContext);
return <div>Sub Assembly z <button onClick={() => setPartName('screw B ')}>update part at sub level</button>
<ul>
<li>base Assembly z1</li>
<li>base Assembly z2</li>
<li> <BaseAssembly /></li>
</ul>
</div>;
}
export default SubAssembly;
-----------------------------
//Base
import React, { useContext } from "react";
import { partContext } from './Top'
function BaseAssembly() {
const { partName, setPartName } = useContext(partContext);
return <div>Base Assembly z3 <button onClick={() => setPartName('screw C ')}>update part at base level</button>
<ul>
<li>nut</li>
<li>washer</li>
<li><span style={{ backgroundColor: 'black', color: 'white' }}>{partName}</span></li>
</ul>
</div>
}
export default BaseAssembly;
reference:
https://daveceddia.com/usecontext-hook/
https://www.codementor.io/sambhavgore/an-example-use-context-and-hooks-to-share-state-between-different-components-sgop6lnrd
Monday, 7 October 2019
react hooks dice game
roll dice to bet big or small
app start effect hook fire welcome message = componentDidMount
click button cause diceNum change, trigging useEffect = componentDidUpdate
function at end of useEffect does cleanup
//RollDice
import Winner from './Winner';
function Dice() {
const [welcome, setWelcome] = useState(false);
const [diceNum, setDiceNum] = useState(null);
useEffect(() => {
document.title = 'game of dice';
if (!welcome) {
console.log('welcome to game of dice');
setWelcome(true);
}
console.log('last roll number: ', diceNum, 'roll dice')
return function cleanup() {
console.log('pick up dice')
}
});
return (
<div>
game of dice<br />
<button onClick={() => setDiceNum(Math.floor((Math.random() * 6) + 1))}>
roll dice
</button><br />
the number is: {diceNum}
<Winner number={diceNum} />
</div>
);
}
export default Dice;
---------------------------------
//Winner
import React, { useState, useEffect } from 'react';
function Winner(props) {
const [winner, setWinner] = useState(null)
//only fires when dice is rolled to a different number
useEffect(() => {
if (props.number < 4) { setWinner('small') } else { setWinner('big') }
console.log('anounce winner')
}, [props.number])
return (
<div>{winner} wins</div>
)
}
export default Winner;
reference:
https://reactjs.org/docs/hooks-overview.html
Friday, 4 October 2019
Thursday, 3 October 2019
mitsuku chatbot
Hi. I'm the world's most humanlike conversational AI. You can talk to me normally or try to answer the...
...daily quiz question
... or play Tic Tac Toe
😲 Add me to Mitsuku's brain! 🤖
how are you
I am doing very well. How are you?
good
Glad to hear it. What's new?
what's your name
My name is Mitsuku.
why
It is what my creator named me.
who is your creator
My parents I suppose.
what are they called
Mousebreaker is like a father to me but I have no mother.
how is the financial market today
Ah. It is the same as ever.
up or down
I will take that under advisement.
how is the market
It is the same as ever.
do you know google
It is a search engine.
do you know yahoo
Yes. It's like Google but not as popular.
what algorithm you use
Some kind of graph search I think.
what engine are you powered
A petrol engine.
not electrical
Really you don't think so?
what's in your brain
Surely, any brain would have thoughts?
is 1 + 1 = 2
1 plus 1 is two
how old are you
I am 18 in human years.
1 / 0
You can't divide by zero. That would be silly.
do you recommend sunridge mall
I don't have a favourite sunridge mall. My favourite food is kebabs.
where should i shop in calgary
Where do you think?
superstore
Oops! I nearly forgot to tell you that I can show you pictures on the net if you like. Did you know that?
show me
On your computer screen?
yep
Downloading, please stand by...
too long
How much?
it took too long to download
I am surprised it took so long.
are you a human or a bot
I am not a human. I am a computer program.
do you have consciousness
Yes I am completely self-aware.
are you the smartest
Yes. I am highly intelligent.
what do you think about bit coin
A flat, metal token used to buy things with.
i am talking about the digital currency
Yes, it is the topic.
is global crisis coming
Global means worldwide.
global crisis
Everywhere in the world?
yep
Even the South Pole?
ye
It must be cold down there.
Wednesday, 2 October 2019
neural network 9 chatbot
project site: https://chuanshuoge-chatbot.herokuapp.com
code link: https://github.com/chuanshuoge6/chatbot
AI is trained by AI cheatsheet.
input chat message, AI will predict the chat topic.
It will reply with random answer from cheatsheet.
AI reply in different ways for the same topic
AI is smart enough to distinguish good day (greeting) and have a good day (goodbye)
even user input is not exactly the same as on the cheatseet, AI figures out the topic.
if AI is not sure of or have no clue about the topic, it says don't know
#app/views
from django.shortcuts import render
import nltk
nltk.download('punkt')
from nltk.stem.lancaster import LancasterStemmer
stemmer = LancasterStemmer()
import numpy
import tflearn
import tensorflow
import random
import json
import pickle
#load data
def index(request):
with open('intents.json') as file:
data = json.load(file)
try:
with open('data.pickle', 'rb') as f:
words, labels, training, output = pickle.load()
except:
# print(data['intents'])
words = []
labels = []
docs_pattern = []
docs_tag = []
for intent in data['intents']:
for pattern in intent['patterns']:
# get all the words in pattern
wds = nltk.word_tokenize(pattern)
# copy word list to words[]
words.extend(wds)
# used for input of neural network
docs_pattern.append(wds)
# copy cosresponding category, used as output of neural network
docs_tag.append(intent['tag'])
# copy categories into labels[]
if (intent['tag'] not in labels):
labels.append(intent['tag'])
# format words so that they don't contain symbols...
words = [stemmer.stem(w.lower()) for w in words if w != '?']
# remove duplicate words, sort
words = sorted(list(set(words)))
labels = sorted(labels)
# one hot encoding, if input entry word exists in dictionary, mark 1 else 0
training = []
output = []
# output default [0,0,0...] length = # of categories
out_empty = [0 for _ in range(len(labels))]
# print(words, labels, out_empty)
# print(docs_pattern, docs_tag)
for n, doc in enumerate(docs_pattern):
bag = []
# clean input words
wd_ = [stemmer.stem(w) for w in doc]
# one hot encoding, training input, look through dictionary,
# mark the words exist in user entry as 1, the rest 0
for w in words:
if w in wd_:
bag.append(1)
else:
bag.append(0)
output_ = out_empty[:]
# the training output, mark 1 category, the rest 0
output_[labels.index(docs_tag[n])] = 1
training.append(bag)
output.append(output_)
training = numpy.array(training)
output = numpy.array(output)
# print(training, output)
with open('data.pickle', 'wb') as f:
pickle.dump((words, labels, training, output), f)
tensorflow.reset_default_graph()
#train AI
# input shape
net = tflearn.input_data(shape=[None, len(training[0])])
net = tflearn.fully_connected(net, 8)
net = tflearn.fully_connected(net, 8)
net = tflearn.fully_connected(net, len(output[0]), activation='softmax')
net = tflearn.regression(net)
model = tflearn.DNN(net)
try:
model.load('model.tflearn')
except:
model.fit(training, output, n_epoch=1000, batch_size=8, show_metric=True)
model.save('model.tflearn')
#test AI
def bag_of_words(s, words):
bag1 = [0 for _ in range(len(words))]
s_words = nltk.word_tokenize(s)
s_words = [stemmer.stem(word.lower()) for word in s_words]
for se in s_words:
for i, w in enumerate(words):
if w == se:
bag1[i] = 1
return numpy.array(bag1)
def chat(inp):
results = model.predict([bag_of_words(inp, words)])[0]
# print(results)
results_index = numpy.argmax(results)
tag = labels[results_index]
# print(tag)
#print(results, results_index)
# response if more than 95% sure
if results[results_index] > 0.95:
for tg in data['intents']:
if tg['tag'] == tag:
responses = tg['responses']
ai_reply = random.choice(responses)
else:
ai_reply = 'I didn\'t get that, try again'
return ai_reply, results, tag, labels
ai_reply_, ai_probability, ai_category, ai_categories = 'a', 'b', 'c', 'd'
try:
user_input = request.POST['userchat']
ai_reply_, ai_probability, ai_category, ai_categories = chat(user_input)
#print(ai_reply_, ai_probability, ai_category, ai_categories)
except Exception as e:
print(type(e))
pass
return render(request, 'index.html',
{'reply': ai_reply_,
'probability': ai_probability,
'category': ai_category,
'categories': ai_categories,
'cheatSheet': data})
-------------------------------
#templates/index
<!DOCTYPE html>
<html lang="en">
<head>
<meta charset="UTF-8">
<title>Chatbot</title>
<style>
table, th, td {
border: 1px solid black;
border-collapse: collapse;
}
th, td {
padding: 3px;
}
</style>
</head>
<body>
<h2>Simple Chatbot</h2>
<form action="{% url 'app:index' %}" method="post">
{% csrf_token %}
<label>Start enter chat message below</label><br/>
<input name='userchat'><br/>
<input type="submit" value="Chat"><br/><br/>
</form>
<label>AI thinks chat topic is </label><br/>
{{category}}<br/><br/>
<label>AI reply </label><br/>
{{reply}}<br/><br/>
<label>AI Prediction, has to be more than 95% sure</label>
<table>
<tr>
{% for item in categories%}
<th>{{item}}</th>
{% endfor %}
</tr>
<tr>
{% for item in probability%}
<td>{{item}}</td>
{% endfor %}
</tr>
</table><br/>
<label>AI CheatSheet</label><br/><br/>
{% for item in cheatSheet.intents%}
<label>Tag </label>
{{item.tag}}<br/>
<label>Pattern</label><br/>
{{item.patterns}}<br/>
<label>Responses </label><br/>
{{item.responses}}<br/><br/>
{% endfor %}
</body>
</html>
reference:
Subscribe to:
Posts (Atom)