[PYTHON] Anchoco for yourself

Due to changes in the site, I don't use it much Forget useful commands ... Since there are many things like that, it is a memo for myself.

Linux

Disk write performance check

 dd bs=1M count=1000 if=/dev/zero of=/HOGE/disktest oflag=direct

wget with cookies

wget -O [output] --load-cookies="./cookie.txt" [url]

GPU status reference

nvidia-smi

Line feed code messing around

perl -p -e 's/\n/\r\n/'

Port confirmation (2017/02/01 added)

lsof -i
lsof -i:[port]
netstat -antu

Python

Set cookie and get

import requests
cookies = {
    'aaa': '81',
    'bbb': 'xxx'
}
res = requests.get([url], cookies=cookies)

List comprehension

http://qiita.com/y__sama/items/a2c458de97c4aa5a98e7

nan judgment

math.iznan(xxx)

No line break print

print("hello", end="")

High speed for

for x in[None]*xxx
for x in[0]*xxx
for x in range(xxx)

Convert to date type (corrected 2017/02/01)

#Easy with pandas
pd.to_datetime('2015/1/1 10:00')

Time zone change (2017/02/01 added)

#Easy with pandas
pd.to_datetime('2015/1/1 10:00').tz_localize('GMT').tz_convert('Asia/Tokyo')
-> Timestamp('2015-01-01 19:00:00+0900', tz='Asia/Tokyo')

Look into modules and classes (Added 2017/02/01)

def pvars(_class):
    for _key in vars(_class).keys():
        print('[%s] %s' % (_key, vars(_class)[_key]))

Find the position of the maximum value in the DataFrame (Added 2017/03/22)

column_name = df.max().argmax()
index_name = df[column_name].argmax()
df.ix[index_name, column_name]

mongodb

Search by type

db.collection.find(xxx: {$type: 2})
    -Missing? 0
    - Double             1
    - String             2
    - Object             3
    - Array              4
    - Null              10
    - 32-bit integer    16
    - Timestamp         17
    - 64-bit integer    18

date type search

db.collection.find({xxx: {$gte:ISODate("2014-11-01T00:00:00Z")}})