性能对比 pypy vs python
性能对比 pypy vs python
不试不知道,一试吓一跳。
如果是CPU密集型的程序,pypy3的执行速度比python要快上一百倍。
talk is cheap, show me the code!
代码很简单,运行加法运算:
执行2千万次
python执行:
python main.py
返回用时:time used 21.422261476516724s
pypy执行:
pypy main.py
返回用时:time used 0.1925642490386963s
差距真的很大。 收起阅读 »
不试不知道,一试吓一跳。
如果是CPU密集型的程序,pypy3的执行速度比python要快上一百倍。
talk is cheap, show me the code!
代码很简单,运行加法运算:
执行2千万次
import time
LOOP = 2*10**8
def add(x,y):
return x+y
def cpu_pressure(loop):
for i in range(loop):
result = add(i,i+1)
if __name__ == '__main__':
start = time.time()
cpu_pressure(LOOP)
print(f'time used {time.time()-start}s')
python执行:
python main.py
返回用时:time used 21.422261476516724s
pypy执行:
pypy main.py
返回用时:time used 0.1925642490386963s
差距真的很大。 收起阅读 »
scrapy源码分析<一>:入口函数以及是如何运行
运行scrapy crawl example 命令的时候,就会执行我们写的爬虫程序。
下面我们从源码分析一下scrapy执行的流程:
执行scrapy crawl 命令时,调用的是Command类
然后我们去看看crawler_process,这个是来自ScrapyCommand,而ScrapyCommand又是CrawlerProcess的子类,而CrawlerProcess又是CrawlerRunner的子类
在CrawlerRunner构造函数里面主要作用就是这个
1. 加载配置文件
默认配置文件defautl_settting.py
load_object的实现
测试代码:
在代码块A中,loader_cls是SpiderLoader,最后返回的的是SpiderLoader.from_settings(settings.frozencopy())
接下来看看SpiderLoader.from_settings,
返回类对象自己,所以直接看__init__函数即可
核心就是这个_load_all_spiders:
走起:
接下来看看_load_spiders
核心就是下面的。
这个obj就是我们平时写的spider类了。
原来分析了这么多,才找到了我们平时写的爬虫类
待续。。。。
原创文章
转载请注明出处
http://30daydo.com/article/530
收起阅读 »
下面我们从源码分析一下scrapy执行的流程:
执行scrapy crawl 命令时,调用的是Command类
class Command(ScrapyCommand):
requires_project = True
def syntax(self):
return '[options]'
def short_desc(self):
return 'Runs all of the spiders - My Defined'
def run(self,args,opts):
print('==================')
print(type(self.crawler_process))
spider_list = self.crawler_process.spiders.list() # 找到爬虫类
for name in spider_list:
print('=================')
print(name)
self.crawler_process.crawl(name,**opts.__dict__)
self.crawler_process.start()
然后我们去看看crawler_process,这个是来自ScrapyCommand,而ScrapyCommand又是CrawlerProcess的子类,而CrawlerProcess又是CrawlerRunner的子类
在CrawlerRunner构造函数里面主要作用就是这个
def __init__(self, settings=None):
if isinstance(settings, dict) or settings is None:
settings = Settings(settings)
self.settings = settings
self.spider_loader = _get_spider_loader(settings) # 构造爬虫
self._crawlers = set()
self._active = set()
self.bootstrap_failed = False
1. 加载配置文件
def _get_spider_loader(settings):
cls_path = settings.get('SPIDER_LOADER_CLASS')
# settings文件没有定义SPIDER_LOADER_CLASS,所以这里获取到的是系统的默认配置文件,
# 默认配置文件在接下来的代码块A
# SPIDER_LOADER_CLASS = 'scrapy.spiderloader.SpiderLoader'
loader_cls = load_object(cls_path)
# 这个函数就是根据路径转为类对象,也就是上面crapy.spiderloader.SpiderLoader 这个
# 字符串变成一个类对象
# 具体的load_object 对象代码见下面代码块B
return loader_cls.from_settings(settings.frozencopy())
默认配置文件defautl_settting.py
# 代码块A
#......省略若干
SCHEDULER = 'scrapy.core.scheduler.Scheduler'
SCHEDULER_DISK_QUEUE = 'scrapy.squeues.PickleLifoDiskQueue'
SCHEDULER_MEMORY_QUEUE = 'scrapy.squeues.LifoMemoryQueue'
SCHEDULER_PRIORITY_QUEUE = 'scrapy.pqueues.ScrapyPriorityQueue'
SPIDER_LOADER_CLASS = 'scrapy.spiderloader.SpiderLoader' 就是这个值
SPIDER_LOADER_WARN_ONLY = False
SPIDER_MIDDLEWARES = {}
load_object的实现
# 代码块B 为了方便,我把异常处理的去除
from importlib import import_module #导入第三方库
def load_object(path):
dot = path.rindex('.')
module, name = path[:dot], path[dot+1:]
# 上面把路径分为基本路径+模块名
mod = import_module(module)
obj = getattr(mod, name)
# 获取模块里面那个值
return obj
测试代码:
In [33]: mod = import_module(module)
In [34]: mod
Out[34]: <module 'scrapy.spiderloader' from '/home/xda/anaconda3/lib/python3.7/site-packages/scrapy/spiderloader.py'>
In [35]: getattr(mod,name)
Out[35]: scrapy.spiderloader.SpiderLoader
In [36]: obj = getattr(mod,name)
In [37]: obj
Out[37]: scrapy.spiderloader.SpiderLoader
In [38]: type(obj)
Out[38]: type
在代码块A中,loader_cls是SpiderLoader,最后返回的的是SpiderLoader.from_settings(settings.frozencopy())
接下来看看SpiderLoader.from_settings,
def from_settings(cls, settings):
return cls(settings)
返回类对象自己,所以直接看__init__函数即可
class SpiderLoader(object):
"""
SpiderLoader is a class which locates and loads spiders
in a Scrapy project.
"""
def __init__(self, settings):
self.spider_modules = settings.getlist('SPIDER_MODULES')
# 获得settting中的模块名字,创建scrapy的时候就默认帮你生成了
# 你可以看看你的settings文件里面的内容就可以找到这个值,是一个list
self.warn_only = settings.getbool('SPIDER_LOADER_WARN_ONLY')
self._spiders = {}
self._found = defaultdict(list)
self._load_all_spiders() # 加载所有爬虫
核心就是这个_load_all_spiders:
走起:
def _load_all_spiders(self):
for name in self.spider_modules:
for module in walk_modules(name): # 这个遍历文件夹里面的文件,然后再转化为类对象,
# 保存到字典:self._spiders = {}
self._load_spiders(module) # 模块变成spider
self._check_name_duplicates() # 去重,如果名字一样就异常
接下来看看_load_spiders
核心就是下面的。
def iter_spider_classes(module):
from scrapy.spiders import Spider
for obj in six.itervalues(vars(module)): # 找到模块里面的变量,然后迭代出来
if inspect.isclass(obj) and \
issubclass(obj, Spider) and \
obj.__module__ == module.__name__ and \
getattr(obj, 'name', None): # 有name属性,继承于Spider
yield obj
这个obj就是我们平时写的spider类了。
原来分析了这么多,才找到了我们平时写的爬虫类
待续。。。。
原创文章
转载请注明出处
http://30daydo.com/article/530
收起阅读 »
crontab定时运行图形程序
默认情况不会显示任何图形的界面,需要在程序前添加
export DISPLAY=:0;
附一个linux下桌面提醒GUI程序,定时提醒你休息哈:
程序保存为task.py
然后设定crontab任务:
* * * * * export DISPLAY=:0; python task.py
即可
收起阅读 »
export DISPLAY=:0;
* * * * * export DISPLAY=:0; gedit
附一个linux下桌面提醒GUI程序,定时提醒你休息哈:
import pyautogui as pag
import datetime
def neck_rest():
f = open('neck_record.txt', 'a')
ret = pag.prompt("Rest! Protect your neck !")
if ret == 'rest':
f.write(datetime.datetime.now().strftime("%Y-%m-%d %H:%M:%S"))
f.write('\t')
f.write('Rest')
f.write('\n')
else:
f.write(datetime.datetime.now().strftime("%Y-%m-%d %H:%M:%S"))
f.write('\t')
f.write('Failed to rest')
f.write('\n')
f.close()
neck_rest()
程序保存为task.py
然后设定crontab任务:
* * * * * export DISPLAY=:0; python task.py
即可
收起阅读 »
python分析目前为止科创板企业省份分布
科创板上市以来已经有一个多月了,我想看看到目前为止,上市企业都是归属哪些地方的。 因为个人觉得科创板是上证板块的,那么来自江浙一带的企业会更多。 毕竟现在深市和沪市在争夺资源,深市希望把深圳企业留回在深市的主板或者中小创版块。
首先获取行情数据,借助tushare这个框架:
在python3环境下,pip install tushare --upgrade ,记得要更新,因为用的旧版本会获取不到科创板的数据。
安装成功后试试import tushare as ts,看看有没有报错。没有就是安装成功了。
接下来抓取全市场的行情.
(点击查看大图)
查看前5条数据
现在行情数据存储在df中,然后分析数据。
因为提取的是全市场的数据,然后获取科创板的企业:
(点击查看大图)
使用的是正则表达式,匹配688开头的代码。
接下来就是分析企业归属地:
(点击查看大图)
使用value_counts函数,统计该列每个值出现的次数。
搞定了! 是不是很简单?
而且企业地区分布和自己的构想也差不多,江浙沪一带占了一半,加上北京地区,占了80%以上的科创板企业了。
每周会定期更新一篇python数据分析股票的文章。
原创文章,欢迎转载
请注明出处:
http://30daydo.com/article/528
收起阅读 »
首先获取行情数据,借助tushare这个框架:
在python3环境下,pip install tushare --upgrade ,记得要更新,因为用的旧版本会获取不到科创板的数据。
安装成功后试试import tushare as ts,看看有没有报错。没有就是安装成功了。
接下来抓取全市场的行情.
(点击查看大图)
查看前5条数据
现在行情数据存储在df中,然后分析数据。
因为提取的是全市场的数据,然后获取科创板的企业:
(点击查看大图)
使用的是正则表达式,匹配688开头的代码。
接下来就是分析企业归属地:
(点击查看大图)
使用value_counts函数,统计该列每个值出现的次数。
搞定了! 是不是很简单?
而且企业地区分布和自己的构想也差不多,江浙沪一带占了一半,加上北京地区,占了80%以上的科创板企业了。
每周会定期更新一篇python数据分析股票的文章。
原创文章,欢迎转载
请注明出处:
http://30daydo.com/article/528
收起阅读 »
python redis.StrictRedis.from_url 连接redis
python redis.StrictRedis.from_url 连接redis
用url的方式连接redis
r=redis.StrictRedis.from_url(url)
url为以下的格式:
原创文章,转载请注明出处:
http://30daydo.com/article/527
收起阅读 »
用url的方式连接redis
r=redis.StrictRedis.from_url(url)
url为以下的格式:
redis://[:password]@localhost:6379/0
rediss://[:password]@localhost:6379/0
unix://[:password]@/path/to/socket.sock?db=0
原创文章,转载请注明出处:
http://30daydo.com/article/527
收起阅读 »
redis health_check_interval 参数无效
因为一直在循环阻塞里面监听redis的发布者,时间长了,redis就掉线了或者网络终端,就会一直卡在等待接受,而发布者后续发布的数据就接收不到了。
而官网的文档说使用参数:
health_check_interval=30 # 30s心跳检测一次
但实际上这个参数在最新的redis 3.3以上是被去掉了。 所以是无办法使用 self.conn = redis.Redis(host='10.18.6.46',health_check_interval=30)
这点在作者的github页面里面也得到了解释。
https://github.com/andymccurdy/redis-py/issues/1199
所以要改成
data = client.blpop('key', timeout=300)
300s后超时,data为None,重新监听。
收起阅读 »
# helper
class RedisHelp(object):
def __init__(self,channel):
# self.pool = redis.ConnectionPool('10.18.6.46',port=6379)
# self.conn = redis.Redis(connection_pool=self.pool)
# 上面的方式无法使用订阅者 发布者模式
self.conn = redis.Redis(host='10.18.6.46')
self.publish_channel = channel
self.subscribe_channel = channel
def publish(self,msg):
self.conn.publish(self.publish_channel,msg) # 1. 渠道名 ,2 信息
def subscribe(self):
self.pub = self.conn.pubsub()
self.pub.subscribe(self.subscribe_channel)
self.pub.parse_response()
print('initial')
return self.pub
helper = RedisHelp('cuiqingcai')
# 订阅者
if sys.argv[1]=='s':
print('in subscribe mode')
pub = helper.subscribe()
while 1:
print('waiting for publish')
pubsub.check_health()
msg = pub.parse_response()
s=str(msg[2],encoding='utf-8')
print(s)
if s=='exit':
break
# 发布者
elif sys.argv[1]=='p':
print('in publish mode')
msg = sys.argv[2]
print(f'msg -> {msg}')
helper.publish(msg)
而官网的文档说使用参数:
health_check_interval=30 # 30s心跳检测一次
但实际上这个参数在最新的redis 3.3以上是被去掉了。 所以是无办法使用 self.conn = redis.Redis(host='10.18.6.46',health_check_interval=30)
这点在作者的github页面里面也得到了解释。
https://github.com/andymccurdy/redis-py/issues/1199
所以要改成
data = client.blpop('key', timeout=300)
300s后超时,data为None,重新监听。
收起阅读 »
mongodb 修改嵌套字典字典的字段名
对于mongodb,修改字段名称的语法是
比如下面的例子:
上面就是把字段corp改为企业。
如果是嵌套字段呢?
比如 corp字典是一个字典,里面是 { 'address':'USA', 'phone':'12345678' }
那么要修改里面的address为地址:
原创文章,转载请注明出处
原文连接:http://30daydo.com/article/521
收起阅读 »
db.test.update({},{$rename:{'旧字段':'新字段'}},true,true)
比如下面的例子:
db.getCollection('example').update({},{$rename:{'corp':'企业'}})
上面就是把字段corp改为企业。
如果是嵌套字段呢?
比如 corp字典是一个字典,里面是 { 'address':'USA', 'phone':'12345678' }
那么要修改里面的address为地址:
db.getCollection('example').update({},{$rename:{'corp.address':'corp.地址'}})
原创文章,转载请注明出处
原文连接:http://30daydo.com/article/521
收起阅读 »
python执行shell命令时报错: -/bin/sh: 命令:not found的解决办法
file='test.txt'
cmd = f'rsync -av {file} root@10.18.6.46:/home/cjw/'
p = subprocess.Popen(cmd, shell=True, stdout=subprocess.PIPE, stderr=subprocess.PIPE,executable="/bin/bash")
output, error = p.communicate()
if p.returncode != 0:
print("Error while running - %s" % cmd)
print(error)
print(output)
用sublime3 运行的时候一直报错。
后来发现,这个是sublime3的运行环境问题, 直接用shell执行 python main.py 执行上面的代码,命令可以正常运行。
/bin/sh: 1: rsync: not found 收起阅读 »
python并行编程手册 勘误
python并行编程手册中文版
65页的进程创建, p.join() 不能写到循环里面,不然的话会阻塞下一次进程的创建,因为下一次进程要卡在join这里。
可以改成这样的
而且后面发现,整本书都是有这个问题的。 收起阅读 »
65页的进程创建, p.join() 不能写到循环里面,不然的话会阻塞下一次进程的创建,因为下一次进程要卡在join这里。
可以改成这样的
p0 = multiprocessing.Process(name=str(0), target=foo, args=(0,))
p0.start()
p1 = multiprocessing.Process(name=str(1), target=foo, args=(1,))
p1.start()
p2 = multiprocessing.Process(name=str(2), target=foo, args=(2,))
p2.start()
p3 = multiprocessing.Process(name=str(3), target=foo, args=(3,))
p3.start()
p4 = multiprocessing.Process(name=str(4), target=foo, args=(4,))
p4.start()
p5 = multiprocessing.Process(name=str(5), target=foo, args=(5,))
p5.start()
p0.join()
p1.join()
p2.join()
p3.join()
p4.join()
p5.join()
而且后面发现,整本书都是有这个问题的。 收起阅读 »
mongodb find得到的数据顺序每次都是一样的
只要用的find内容不变,那么返回的内容顺序也就都一样的。
[Articles to save]
Since on Raspberrypi and can't launch note application , using this web page to save articles link to store later.
https://www.jisilu.cn/question/321759 -Done
https://www.80shihua.com/archives/1590 -Done
收起阅读 »
https://www.jisilu.cn/question/321759 -Done
https://www.80shihua.com/archives/1590 -Done
收起阅读 »
Raspberrypi 2 Install or upgrade Python3.6
Since no chinese input method in my raspberrypi, i can only write with English.
Raspberrypi has python2. 7 and python3.4, but i want to upgrade to python3.6+.
Python3.6 support some new feature such as print(f'{name}') and x=1_000_242_200 expression.
How to upgrade ?
then run command:
wait for about 20mins (low perf of raspberrypi :( )
then you run command:
python3
it will using the new python3.6 version:
Enjoy it ! 收起阅读 »
Raspberrypi has python2. 7 and python3.4, but i want to upgrade to python3.6+.
Python3.6 support some new feature such as print(f'{name}') and x=1_000_242_200 expression.
How to upgrade ?
$ wget https://www.python.org/ftp/pyt ... 1.tgz
$ tar zxvf Python-3.6.1.tgz $ cd Python-3.6.1
then run command:
$ sudo ./configure && sudo make && sudo make install
wait for about 20mins (low perf of raspberrypi :( )
then you run command:
python3
it will using the new python3.6 version:
Python 3.6.1 (default, Jul 21 2019, 14:26:28)
[GCC 4.9.2] on linux
Type "help", "copyright", "credits" or "license" for more information.
Enjoy it ! 收起阅读 »
frontera运行link_follower.py 报错:doesn't define any object named 'FIFO'
代码如下:
无论用的py2或者py3,都会报以下的错误。
from __future__ import print_function
import re
import requests
from frontera.contrib.requests.manager import RequestsFrontierManager
# from frontera.contrib.requests.manager import RequestsFrontierManager
from frontera import Settings
from six.moves.urllib.parse import urljoin
SETTINGS = Settings()
SETTINGS.BACKEND = 'frontera.contrib.backends.memory.FIFO'
# SETTINGS.BACKEND = 'frontera.contrib.backends.memory.MemoryDistributedBackend'
SETTINGS.LOGGING_MANAGER_ENABLED = True
SETTINGS.LOGGING_BACKEND_ENABLED = True
SETTINGS.MAX_REQUESTS = 100
SETTINGS.MAX_NEXT_REQUESTS = 10
SEEDS = [
'http://www.imdb.com',
]
LINK_RE = re.compile(r'<a.+?href="(.*?)".?>', re.I)
def extract_page_links(response):
return [urljoin(response.url, link) for link in LINK_RE.findall(response.text)]
if __name__ == '__main__':
frontier = RequestsFrontierManager(SETTINGS)
frontier.add_seeds([requests.Request(url=url) for url in SEEDS])
while True:
next_requests = frontier.get_next_requests()
if not next_requests:
break
for request in next_requests:
try:
response = requests.get(request.url)
links = [
requests.Request(url=url)
for url in extract_page_links(response)
]
frontier.page_crawled(response)
print('Crawled', response.url, '(found', len(links), 'urls)')
if links:
frontier.links_extracted(request, links)
except requests.RequestException as e:
error_code = type(e).__name__
frontier.request_error(request, error_code)
print('Failed to process request', request.url, 'Error:', e)
无论用的py2或者py3,都会报以下的错误。
raise NameError("Module '%s' doesn't define any object named '%s'" % (module, name))收起阅读 »
NameError: Module 'frontera.contrib.backends.memory' doesn't define any object named 'FIFO'
scrapy-rabbitmq 不支持python3 [修改源码使它支持]
官方版本在2015年就没有更新了。
在python3上运行的收会报错。
需要修改以下地方:
待续。。
在python3上运行的收会报错。
需要修改以下地方:
待续。。
scrapy rabbitmq 分布式爬虫
对于没接触过rabbitmq的同学,可以看这个文章:https://blog.csdn.net/hellozpc/article/details/81436980
rabbitmq是个不错的消息队列服务,可以配合scrapy作为消息队列.
下面是一个简单的demo:
启动spider:
然后往rabbitmq里面推送数据:
推送数据后,scrapy会马上接受到队里里面的数据。
注意不能在start_requests里面写等待队列的命令,因为start_requests函数需要返回一个生成器,否则程序会报错。
待续。。。
###### 2019-08-29 更新 ###################
发现一个坑,就是rabbitMQ在接受到数据后,无法在回调函数里面使用yield生成器。
收起阅读 »
rabbitmq是个不错的消息队列服务,可以配合scrapy作为消息队列.
下面是一个简单的demo:
import re
import requests
import scrapy
from scrapy import Request
from rabbit_spider import settings
from scrapy.log import logger
import json
from rabbit_spider.items import RabbitSpiderItem
import datetime
from scrapy.selector import Selector
import pika
# from scrapy_rabbitmq.spiders import RabbitMQMixin
# from scrapy.contrib.spiders import CrawlSpider
class Website(scrapy.Spider):
name = "rabbit"
def start_requests(self):
headers = {'Accept': '*/*',
'Accept-Encoding': 'gzip, deflate, br',
'Accept-Language': 'en-US,en;q=0.9,zh-CN;q=0.8,zh;q=0.7',
'Host': '36kr.com',
'Referer': 'https://36kr.com/information/web_news',
'User-Agent': 'Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/74.0.3729.108 Safari/537.36'
}
url = 'https://36kr.com/information/web_news'
yield Request(url=url,
headers=headers)
def parse(self, response):
credentials = pika.PlainCredentials('admin', 'admin')
connection = pika.BlockingConnection(pika.ConnectionParameters('192.168.1.101', 5672, '/', credentials))
channel = connection.channel()
channel.exchange_declare(exchange='direct_log', exchange_type='direct')
result = channel.queue_declare(exclusive=True, queue='')
queue_name = result.method.queue
# print(queue_name)
# infos = sys.argv[1:] if len(sys.argv)>1 else ['info']
info = 'info'
# 绑定多个值
channel.queue_bind(
exchange='direct_log',
routing_key=info,
queue=queue_name
)
print('start to receive [{}]'.format(info))
channel.basic_consume(
on_message_callback=self.callback_func,
queue=queue_name,
auto_ack=True,
)
channel.start_consuming()
def callback_func(self, ch, method, properties, body):
print(body)
启动spider:
from scrapy import cmdline
cmdline.execute('scrapy crawl rabbit'.split())
然后往rabbitmq里面推送数据:
import pika
import settings
credentials = pika.PlainCredentials('admin','admin')
connection = pika.BlockingConnection(pika.ConnectionParameters('192.168.1.101',5672,'/',credentials))
channel = connection.channel()
channel.exchange_declare(exchange='direct_log',exchange_type='direct') # fanout 就是组播
routing_key = 'info'
message='https://36kr.com/pp/api/aggregation-entity?type=web_latest_article&b_id=59499&per_page=30'
channel.basic_publish(
exchange='direct_log',
routing_key=routing_key,
body=message
)
print('sending message {}'.format(message))
connection.close()
推送数据后,scrapy会马上接受到队里里面的数据。
注意不能在start_requests里面写等待队列的命令,因为start_requests函数需要返回一个生成器,否则程序会报错。
待续。。。
###### 2019-08-29 更新 ###################
发现一个坑,就是rabbitMQ在接受到数据后,无法在回调函数里面使用yield生成器。
收起阅读 »
exchange_declare() got an unexpected keyword argument 'type'
In new version of pika, now it is using
exchange_type instead of type
exchange_type instead of type
credentials = pika.PlainCredentials('admin','admin')收起阅读 »
connection = pika.BlockingConnection(pika.ConnectionParameters('192.168.1.101',5672,'/',credentials))
channel = connection.channel()
channel.exchange_declare(exchange='logs',exchange_type='fanout')
twisted的getPage已经不建议使用,新接口为twisted.web.client.Agent
Twisted-16.7.0 is coming soon, and it deprecates twisted.web.client.getPage (and client.HTTPClientFactory). We use these in some of the unit tests, to fetch one of the HTTP WAPI/WUI pages and make sure the contents look right.
We need to change these tests to use twisted.web.client.Agent instead, or a package named "treq", which is a Twisted flavor of the excellent (but blocking) requests library.
收起阅读 »
twisted reactor运行后,添加了addBoth函数,但是还是无法停止
代码如下:
上面的代码是无法停止的,如果使用的是
dd.addBoth(done)
done函数的定义是没有参数的。
而使用另一个done函数带参数的done(*args,**kwargs)
是可以正常退出的,done里面写了reactor.stop() 函数
原创文章
转载请注明出处:
http://30daydo.com/article/509
收起阅读 »
from scrapy.selector import Selector
def get_response_callback(content):
txt = str(content,encoding='utf-8')
resp = Selector(text=txt)
title = resp.xpath('//title/text()').extract_first()
print(title)
@defer.inlineCallbacks
def task():
url = 'http://www.baidu.com'
d=getPage(url.encode('utf-8'))
d.addCallback(get_response_callback)
yield d
def done():
reactor.stop()
def done1(*args,**kwargs):
reactor.stop()
task_list =
for i in range(4):
d=task()
task_list.append(d)
dd = defer.DeferredList(task_list)
dd.addBoth(done)
reactor.run()
上面的代码是无法停止的,如果使用的是
dd.addBoth(done)
done函数的定义是没有参数的。
而使用另一个done函数带参数的done(*args,**kwargs)
是可以正常退出的,done里面写了reactor.stop() 函数
原创文章
转载请注明出处:
http://30daydo.com/article/509
收起阅读 »
numpy indices的用法
Suppose you have a matrix M whose (i,j)-th element equals
M_ij = 2*i + 3*j
One way to define this matrix would be
i, j = np.indices((2,3))
M = 2*i + 3*j
which yields
array([[0, 3, 6],
[2, 5, 8]])
In other words, np.indices returns arrays which can be used as indices. The elements in i indicate the row index:
In [12]: i
Out[12]:
array([[0, 0, 0],
[1, 1, 1]])
The elements in j indicate the column index:
In [13]: j
Out[13]:
array([[0, 1, 2],
[0, 1, 2]])
上面是Stack Overflow的解释。 翻译一下:
np.indices((2,3))
返回的是一个行列的索引,然后可以用这个索引快速的创建二维数据。
比如我要画一个圆:
img = np.zeros((400,400))
ir,ic = np.indices(img.shape)
circle = (ir-135)**2+(ic-150)**2 < 30**2 # 半径30,圆心在135,150
img[circle]=1
img现在就是一个圆啦
收起阅读 »
cv2 distanceTransform函数的用法 python
distanceTransform
Calculates the distance to the closest zero pixel for each pixel of the source image.
Python: cv2.distanceTransform(src, distanceType, maskSize[, dst]) → dst
Python: cv.DistTransform(src, dst, distance_type=CV_DIST_L2, mask_size=3, mask=None, labels=None) → None
Parameters:
src – 8-bit, single-channel (binary) source image.
dst – Output image with calculated distances. It is a 32-bit floating-point, single-channel image of the same size as src .
distanceType – Type of distance. It can be CV_DIST_L1, CV_DIST_L2 , or CV_DIST_C .
maskSize – Size of the distance transform mask. It can be 3, 5, or CV_DIST_MASK_PRECISE (the latter option is only supported by the first function). In case of the CV_DIST_L1 or CV_DIST_C distance type, the parameter is forced to 3 because a 3\times 3 mask gives the same result as 5\times 5 or any larger aperture.
labels – Optional output 2D array of labels (the discrete Voronoi diagram). It has the type CV_32SC1 and the same size as src . See the details below.
labelType – Type of the label array to build. If labelType==DIST_LABEL_CCOMP then each connected component of zeros in src (as well as all the non-zero pixels closest to the connected component) will be assigned the same label. If labelType==DIST_LABEL_PIXEL then each zero pixel (and all the non-zero pixels closest to it) gets its own label.
The functions distanceTransform calculate the approximate or precise distance from every binary image pixel to the nearest zero pixel. For zero image pixels, the distance will obviously be zero.
When maskSize == CV_DIST_MASK_PRECISE and distanceType == CV_DIST_L2 , the function runs the algorithm described in [Felzenszwalb04]. This algorithm is parallelized with the TBB library.
In other cases, the algorithm [Borgefors86] is used. This means that for a pixel the function finds the shortest path to the nearest zero pixel consisting of basic shifts: horizontal, vertical, diagonal, or knight’s move (the latest is available for a 5\times 5 mask). The overall distance is calculated as a sum of these basic distances. Since the distance function should be symmetric, all of the horizontal and vertical shifts must have the same cost (denoted as a ), all the diagonal shifts must have the same cost (denoted as b ), and all knight’s moves must have the same cost (denoted as c ). For the CV_DIST_C and CV_DIST_L1 types, the distance is calculated precisely, whereas for CV_DIST_L2 (Euclidean distance) the distance can be calculated only with a relative error (a 5\times 5 mask gives more accurate results). For a,``b`` , and c , OpenCV uses the values suggested in the original paper:
CV_DIST_C (3\times 3) a = 1, b = 1
CV_DIST_L1 (3\times 3) a = 1, b = 2
CV_DIST_L2 (3\times 3) a=0.955, b=1.3693
CV_DIST_L2 (5\times 5) a=1, b=1.4, c=2.1969
Typically, for a fast, coarse distance estimation CV_DIST_L2, a 3\times 3 mask is used. For a more accurate distance estimation CV_DIST_L2 , a 5\times 5 mask or the precise algorithm is used. Note that both the precise and the approximate algorithms are linear on the number of pixels.
The second variant of the function does not only compute the minimum distance for each pixel (x, y) but also identifies the nearest connected component consisting of zero pixels (labelType==DIST_LABEL_CCOMP) or the nearest zero pixel (labelType==DIST_LABEL_PIXEL). Index of the component/pixel is stored in \texttt{labels}(x, y) . When labelType==DIST_LABEL_CCOMP, the function automatically finds connected components of zero pixels in the input image and marks them with distinct labels. When labelType==DIST_LABEL_CCOMP, the function scans through the input image and marks all the zero pixels with distinct labels.
In this mode, the complexity is still linear. That is, the function provides a very fast way to compute the Voronoi diagram for a binary image. Currently, the second variant can use only the approximate distance transform algorithm, i.e. maskSize=CV_DIST_MASK_PRECISE is not supported yet.
Note
An example on using the distance transform can be found at opencv_source_code/samples/cpp/distrans.cpp
(Python) An example on using the distance transform can be found at opencv_source/samples/python2/distrans.py
收起阅读 »
Win10下PhantomJS无法运行 【版本兼容问题】
以前在win7上运行的好好的。
在win10下就报错:
selenium.common.exceptions.WebDriverException: Message: Service C:\Tool\phantomjs-2.5.0-beta2-windows\phantomjs-2.5.0-beta2-windows\bin\phantomjs.exe unexpectedly exited. Status code was: 4294967295
后来替换了一个旧的版本,发现问题就这么解决了。
旧版本:phantomjs-2.1.1-windows
原创文章
转载请注明出处
http://30daydo.com/article/505
收起阅读 »
在win10下就报错:
selenium.common.exceptions.WebDriverException: Message: Service C:\Tool\phantomjs-2.5.0-beta2-windows\phantomjs-2.5.0-beta2-windows\bin\phantomjs.exe unexpectedly exited. Status code was: 4294967295
后来替换了一个旧的版本,发现问题就这么解决了。
旧版本:phantomjs-2.1.1-windows
原创文章
转载请注明出处
http://30daydo.com/article/505
收起阅读 »
nunpy中的std标准差是样本差吗
写个代码测试下:
5.5
X.std() # 标准差
2.8722813232690143
手工计算一下:
result = my_fangca(X)
得到的result
(8.25, 2.8722813232690143)
说明numpy的std是标准差,不是样本差 收起阅读 »
# 测试一下那个方差X.mean()
x=[1,2,3,4,5,6,7,8,9,10]
X = np.array(x)
5.5
X.std() # 标准差
2.8722813232690143
手工计算一下:
def my_fangca(X):
l=len(X)
mean=X.mean()
sum_ = 0
sum_std=0
for i in X:
sum_+=(i-mean)**2
var_=sum_/l
std_=(sum_/(l))**0.5
return var_,std_
result = my_fangca(X)
得到的result
(8.25, 2.8722813232690143)
说明numpy的std是标准差,不是样本差 收起阅读 »