eerywh的博客

【代码】jupyter lab环境配置。

pip 虚拟环境、代理、conda虚拟环境

https://www.perceptualedge.com/blog/?p=2080

!pip install sort-dataframeby-monthorweek!pip install sorted-months-weekdayshttps://www.cnpython.com/pypi/sort-dataframeby-monthorweek按MonthorWeek排序数据框架

pandas 中get_dummies() 与factorize()的区别当一个特征中存在较多的类别时，使用get_dummies() 会导致DataFrame中的columns 列数激增factorize() 可以对特征中的类别创建一些数字，来表示分类变量或者枚举型变量(enumerated type)。具体来说：factorize() 只产生一个特征变量，这个特征中对类别使用数字进行区分。适用于出现较多独立变量的情况参考e.p....

import numpy as npimport pandas as pdimport matplotlib.pyplot as pltimport matplotlib.image as mpimg%matplotlib inlineimport seaborn as snsfrom pprint import pprintRead Dataairbnb = pd.read_csv('./new-york-city-airbnb-open-data/AB_NYC_2019.csv')

解决方法：conda clean -i

两个变量之间的相关性检验1、皮尔森相关系数：它们之间的协方差cov(X,Y)除以它们各自标准差的乘积(σX, σY)。要判断两个变量之间是否存在相关性：可以采用画散点图的方式，进行直观性的分析。使用python进行person检验import pandas as pdimport numpy as npimport missingnodf = pd.read_csv('./titanic/train.csv')df.head() P

用Python进行全面的数据探索import pandas as pdimport numpy as npfrom matplotlib import pyplot as pltfrom scipy.stats import normfrom sklearn.preprocessing import StandardScalerfrom scipy import statsimport missingno as mnoimport seaborn as snsimport warnings

conda 导出安装包列表conda list -e > condalist.txt导入conda install --yes --file condalist.txt

用于计算MAC地址匹配数量import pandas as pdimport numpy as npimport reimport ospath = 'D:\Personal\Downloads'df1 = pd.read_csv(os.path.join(path,'1.csv'),skiprows = 7 ,encoding='gbk')df2 = pd.read_csv(os.path.join(path,'2.csv'),header = None)df2.rename(colu

偏度和峰度偏度（skewness）数据分布的的不对称性称为偏态。偏态是指数据分布的偏斜方向和程度。偏度，通常分为右偏（正偏）和左偏（负偏）两种。它们是以对称分布为标准相比较而言的。在对称分布的情况下，平均数、中位数和众数是合而为一的。在偏态分布的情况下，平均数、中位数与众数是分离的。如果众数在左边，平均数在右边，即数据的极端值在右边，数据分布曲线向右延伸，则称为右向偏态。右向偏态，众数的数值较小，平均数的数值较大，平均数与众数之差为正值，因此右向偏态又称为正向偏态。如果众数在右边，平均数在左边，

missingnmissingn --github# 安装pip install missingno快速开始pip install quiltquilt install ResidentMario/missingno_data学习地址：https://www.cnblogs.com/yongestcat/p/11458057.html

共享单车EDA与模型选择import pandas as pd import numpy as npfrom matplotlib import pyplot as pltimport seaborn as snsfrom datetime import datetimeimport osimport warningswarnings.filterwarnings(action = 'ignore')Kaggle-competition-bike-sharing-demandEDA

由于国内使用git clone 下载速度太慢，考虑使用代理取消代理git config --global --unset http.proxy git config --global --unset https.proxy根据代理开的端口设置，我的为10808设置代理git config --global http.proxy socks5://127.0.0.1:10808git config --global https.proxy socks5://127.0.0.1:10808gi

记录在正则表达式中运用到的网站正则表达式简介 正则表达式在线测试提供在线测试，及一些常用的正则表达式语法样例。

一个简单的 Jupyter notebook 输出中文PDF文件的方法直接浏览器虚拟打印 ，其实看着还是不错的！

import requestsimport numpyimport reimport time# 加载pymongoimport pymongoclient = pymongo.MongoClient('localhost',27017)book_weather = client['weather']sheet_weather = book_weather['sheet_weat...

Altair的使用学习# !pip install vega# !pip install vega_datasetsimport altair as altalt.renderers.enable('notebook')RendererRegistry.enable('notebook')# load a simple dataset as a pandas DataFrame...

# data from http://grouplens.org/datasets/movielens/# import os# data_folder = os.path.join(os.path.expanduser("~"), "Data", "ml-100k")# ratings_filename = os.path.join(data_folder, "u.data")fro...

Data CleansingCleaning Rows with NaNsimport pandas as pddf = pd.read_csv('NaNDataset.csv')df.isnull().sum()A 0B 2C 0dtype: int64print(df) A B C0 1 2.0 31 4 ...

#/boot/grub/grub.conf 缺失： yum install -y grubgrub-mkconfig -o /boot/grub/grub.conf #/boot/grub2/grub.cfg 缺失： yum install -y grub2grub2-mkconfig -o /boot/grub2/grub.cfg

IntroductionThis is a quick way to look at the different Datasets…import numpy as np # linear algebraimport pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)# First, look at everything.from subprocess import check_output # 接受上一条命

需要安装的包 : torch ,torchvision,easyocrtorch,torchvision 对应版本表python 包下载地址推荐torchvision包下载地址pip install torchvision-0.8.2+cpu-cp38-cp38-win_amd64.whlpip install torch-1.7.1-cp38-cp38-win_amd64.whlpip install easyocr-1.2.1-py3-none-any.whl...

pd.to_datetimec_itcont['TRANSACTION_DT'].head()0 72420201 72420202 73120203 81020204 8072020Name: TRANSACTION_DT, dtype: int64c_itcont['TRANSACTION_DT'] = c_itcont['TRANSACTION_DT'].astype('str')c_itcont['TRANSACTION_DT'] = pd.to_

Tipsfrom tabulate import tabulate

01.05 shell commandreferencesosx:~ $ echo "hello world" # echo is like Python's print functionhello worldosx:~ $ pwd # pwd = print working directory/home/jake # this is the "path" tha

import pandas as pdimport numpy as np使用matplotlib 画图# To plot pretty figures directly within Jupyter%matplotlib inlineimport matplotlib as mplmpl.rc('axes', labelsize=14)mpl.rc('xtick', labelsize=12)mpl.rc('ytick', labelsize=12).loc selects data

pandas.get_dummiespandas 提供的生成独热编码的方法pandas.get_dummies(data, prefix=None, prefix_sep=’_’, dummy_na=False, columns=None, sparse=False, drop_first=False, dtype=None)[source]pandas.get_dummiesimport pandas as pddf = pd.DataFrame([ ['green'

查看Python版本# cmd python --version查看Python 安装的所有的包pip list*查看Python安装的某个包的版本pip show scipy

pd 列显示百分比（精确到两位小数）```pythondf_total['win10比例'] = (df_total['win10数量'] / df_total['计算机总数']).apply(lambda x: '%.2f%%' % (x*100))df.to_csv 中文显示乱码# df_total.to_csv('11.csv',encoding='utf_8')df_total.to_csv('11.csv',encoding='utf_8_sig')...

pd.loc vs pd.iloc 参考# importing pandas and numpyimport pandas as pdimport numpy as np# crete a sample dataframedata = pd.DataFrame({ 'age' : [ 10, 22, 13, 21, 12, 11, 17], 'section' : [ 'A', 'B', 'C', 'B', 'B', 'A', 'A'], 'city' :

Anaconda 更新到最新版打开需要更新环境的Anaconda promptconda update conda conda update anaconda