Kp0fS的草稿纸

pip install openpyxledata = pd.read_excel(excel)

# 这个脚本是将r'Y:\Project\SCD\SDVDS\VFDVD-Rename'下的d号文件名，重命名为'2021-04-07--0031-P1'rootPath = r'Y:\Project\VDFVD\rVDF00DFVD030\VV-Rename'excel = r'Y:\Project\VDFVD\r00VFDV030\VDFVF数据.xlsx'import osimport pandas as pdlsDir = os.listdir(rootPath)edata =.

import numpy as npimport pandas as pdimport matplotlib.pyplot as pltdef getData(): r""" 使用numpy构造满足条件的随机数() :return: """ dataX = np.random.randint(1,100,size=(1,100)) dataY = np.random.randint(1,100,size=(1,100)) data = ...

方案1https://www.jianshu.com/p/5cf2abe4385c方案2：https://www.pythonf.cn/read/91137方案3：x_train, x_test, y_train, y_test = train_test_split(data, label, test_size=0.15, ramdon_state=0)

介绍的还是挺详细的。

Before = np.arange(range[0]-frame*n, range[0], n)During = np.arange(range[0], range[1], n)After = np.arange(range[1], range[1]+frame*n, n)Total = np.hstack((Before, During, After))

为什么pycharm的jupyter的debug没有 debug python consle（用起来真不方便啊）坑已经反馈给anna了。希望他们能早日解决。

https://blog.youkuaiyun.com/u011412768/article/details/109522349这篇博客总结的非常仔细。非常好！

newdf = pd.DataFrame(data = df.value.T)

https://blog.youkuaiyun.com/weixin_43828405/article/details/92801613推荐使用subplots

N09_0027_P1 = N09_0027_P1.reset_index(drop=True)

https://matplotlib.org/stable/tutorials/intermediate/legend_guide.html#sphx-glr-tutorials-intermediate-legend-guide-py

N09_0027_P1['mean pupil'].plot(xlabel='date', ylabel='pupil size', kind='line', title='N09_0027_P1',ax = plt.subplot(411))N09_0027_P2['mean pupil'].plot(xlabel='date', ylabel='pupil size', kind='line', title='N09_0027_P2',ax = plt.subplot(412))N09_0027_.

https://pandas.pydata.org/pandas-docs/stable/reference/api/pandas.DataFrame.plot.html请看如上的文档plt.figure()N30_0032_P4['mean pupil'].plot(xlabel='date',ylabel='pupil size',kind = 'line', title = 'N30_0032_P4')plt.show()N30_0032_P3['mean pupil'].pl.

N09_P1 = dfResults.loc[dfResults['des'].isin(['0027-P1'])] #绝对查询found = dfResults.loc[dfResults['des'].str.contains('0027-P1')]#模糊查询

search = ['python','java','go'] found = [s['language].str.contains(x) for x in searchfor]

https://www.cnblogs.com/zmc940317/p/13615168.html这个博客总结的非常好。# 选取某列是否是某一类型的数值 用 isindf.loc[df['column_name'].isin(some_values)]

Python pandas to csv 如何保存指定列 （根据列名保存csv）解决方案：一定要有两个列表哦

pandas 如何计算两个 series 之间的相关性how to calculate the correlation bewteen twoseriesgoogle yyds

比方说我有一个需求：新添加一列？---那么用的应该是axis = 1dfData2['PupilSize'] = dfData2.apply(calPupilSize, axis=1)这样理解：新添一列，那么之前所有列都参与计算了。或者，新添加一行？ axis = 0NaCl0_9dfDataMean = NaCl0_9dfData.mean(axis=0)上边的内容太难记忆了，怎么想一个好的比喻呢？看0，相当于把所有的行给圈住了。比如...

直接调用pd的函数，然后指定行、列

NaCl12dfData5 = pd.read_csv(NaCl12FilePath,index_col = 0)在csv中 第一列是index如果csv中没有列名，该怎么办？header = None

https://www.jb51.net/article/137888.htm还有其他若干方法。

坑index=False

label = 'line'l1, = plt.plot(x, y1, label = 'line')plot的时候设置lablel2, = plt.plot(x, y2, label = 'parabola', color = 'red', linewidth = 1.0, linestyle = '--')

plot.line 加个title， 加个图标legendbodyPart1['dis'].plot.line(style = 'r')bodyPart2['dis'].plot.line(style = 'g')bodyPart3['dis'].plot.line(style = 'b')ax = bodyPart4['dis'].plot.line()ax.set_xlabel('Frame')plt.legend(loc='best')plt.show()...

help(pd.DataFrame.plot)lines = bodyPart1['dis'].plot.line(x='pig', y='horse')参考：https://blog.youkuaiyun.com/h_hxx/article/details/90635650我的问题，（1）如何使用？（2）如何有效记住？df.plot.line()参考：https://blog.youkuaiyun.com/h_hxx/article/details/9...

pandas每一列都执行相同的函数apply函数。我的需求：（1）第三列，是第一列和第二列的平方和，开根号。df['TotalScore'] = df['Score'] + df['ExtraScore'] #只能用加法...因为加法运算符被重载过了。好像这种句式就可以。不行！！！必须得用apply函数def calDistance(df): print(df) a = bodyPart1['x'] b = ...

用excel打开csv，把中文字符删了。然后另存为csv-utf的格式。就好了

oneVector2 = pd.DataFrame(data =np.random.random((1,3)))oneVector1 = pd.DataFrame(data =np.random.random((1,3)))按照“行”进行拼接：ccc = pd.concat([oneVector1,oneVector2],axis =0,ignore_index = True)如何对列进行拼接呢？ccc = pd.concat([oneVector1,...

timeStr = time.strftime("%Y-%m-%d-%H-%M-%S", time.localtime()) + "-" + PMStr + "-Report.csv"dfDate3.to_csv(timeStr,mode="w")

r.values[0]

结论是用numpy，然后重新转换成pandas

X.T.dot(X)

df.astype("int")

BI是Business Intelligence(BI)数据分析一般分为三个层面：（1）Memory内存分析层面（2）BI分析层面（3）Massive分析层面

Python 绘制直方图plt.hist(h)plt.show()rNumpyprint('axis sum',rNumpy.sum(axis=1))#每个客户一共用的代数h = rNumpy.sum(axis=1)num_bins = 300#一共135个人plt.hist(h,num_bins, facecolor='blue', alpha=0.5)plt.show()#只统计1~100的...