pandas对csv文件的相关操作

快来打我* 2023-06-19 06:53 130阅读 0赞

# pandas csv文件相关函数和相关操作 #

**0 注意:inplace = True表示当前操作已经在应用到当前环境中了,比如删除一行,然后inplace = True之后,下次再使用就没有删除的这一行了  
1 np.set\_printoptions(threshold=np.inf) numpy时使用最大数显示  
2 data = pd.read\_csv(r"data.csv",nrows=5) 加载固定行数数据  
3 data.head(2) 显示固定行数  
4 data\[“index”\].values\[0:10\] \#标签为index的列数据  
5 data.columns.values 显示所有标签  
6 data.drop(data.index\[1\],inplace=True) 删除第一行  
7 data1.drop\_duplicates(‘name’, inplace=True) 删除"name"中的重复数据  
8 data.rename(columns=\{‘index-given’:‘u7777777’,‘wgsize’:‘wwww’\}, inplace = True) 替换列表列标签  
9 data.insert(5,‘typeSmallSmall’,insert1) \#在第五个空子(0,1,2,3,4,5)插入列名字为typeSmallSmall的数据insert1**

import pandas as pd
    import numpy as np
    np.set_printoptions(threshold=np.inf) #最大数显示

# data = pd.read_csv(r"data.csv",nrows=5) #加载固定行数
    data = pd.read_csv(r"data.csv")
    # data

data.head(2) #显示前五行

<table> 
 <thead> 
  <tr> 
   <th></th> 
   <th>index</th> 
   <th>index-given</th> 
   <th>benchmark</th> 
   <th>dataset</th> 
   <th>comp</th> 
   <th>rational</th> 
   <th>mem</th> 
   <th>localmem</th> 
   <th>coalesced</th> 
   <th>atomic</th> 
   <th>...</th> 
   <th>wgsize</th> 
   <th>transfer1</th> 
   <th>wgsize1</th> 
   <th>transfer.1</th> 
   <th>wgsize.1</th> 
   <th>oracle</th> 
   <th>runtime_cpu</th> 
   <th>runtime_gpu</th> 
   <th>src</th> 
   <th>seq</th> 
  </tr> 
 </thead> 
 <tbody> 
  <tr> 
   <td>0</td> 
   <td>0</td> 
   <td>0</td> 
   <td>amd-app-sdk-3.0-BinomialOption-binomial_options</td> 
   <td>default</td> 
   <td>98</td> 
   <td>8</td> 
   <td>13</td> 
   <td>11</td> 
   <td>2</td> 
   <td>0</td> 
   <td>...</td> 
   <td>255</td> 
   <td>0.000</td> 
   <td>0.249</td> 
   <td>2048</td> 
   <td>255</td> 
   <td>GPU</td> 
   <td>3.291073</td> 
   <td>0.152038</td> 
   <td>__kernel void A(int a, const __global float4* ...</td> 
   <td>[129 129 129 ..., 0 127 0]</td> 
  </tr> 
  <tr> 
   <td>1</td> 
   <td>1</td> 
   <td>1</td> 
   <td>amd-app-sdk-3.0-BitonicSort-bitonicSort</td> 
   <td>default</td> 
   <td>12</td> 
   <td>2</td> 
   <td>6</td> 
   <td>0</td> 
   <td>6</td> 
   <td>0</td> 
   <td>...</td> 
   <td>256</td> 
   <td>0.288</td> 
   <td>0.250</td> 
   <td>131072</td> 
   <td>256</td> 
   <td>GPU</td> 
   <td>0.121940</td> 
   <td>0.057834</td> 
   <td>__kernel void A(__global uint* a, const uint b...</td> 
   <td>[129 129 129 ..., 0 127 0]</td> 
  </tr> 
 </tbody> 
</table>

2 rows × 21 columns

print(data["index"].values[0:10]) #标签为index的列数据

[0 1 2 3 4 5 6 7 8 9]

print(data.columns.values) #第一行,所有标签

['index' 'index-given' 'benchmark' 'dataset' 'comp' 'rational' 'mem'
     'localmem' 'coalesced' 'atomic' 'transfer' 'wgsize' 'transfer1' 'wgsize1'
     'transfer.1' 'wgsize.1' 'oracle' 'runtime_cpu' 'runtime_gpu' 'src' 'seq']

print(data["index-given"].index[1])
    data.drop(data.index[1],inplace=True) #直接删除csv的数据中的第某一行行
    print(data["index-given"].index[1])

1
    2

data1 = pd.read_csv(r"data1.csv")
    print(len(data1["name"]))
    data1.drop_duplicates('name', inplace=True) #删除"name"中的重复数据
    print(len(data1["name"]))
    data1.to_csv(r"data2.csv")

25
    20

print(data.columns.values)
    data.drop(['benchmark'],axis=1,inplace=True) #删除列 index
    print(data.columns.values)

['index-given' 'benchmark' 'dataset' 'comp' 'rational' 'mem' 'localmem'
     'coalesced' 'atomic' 'transfer' 'wgsize' 'transfer1' 'wgsize1'
     'transfer.1' 'wgsize.1' 'oracle' 'runtime_cpu' 'runtime_gpu' 'src' 'seq']
    ['index-given' 'dataset' 'comp' 'rational' 'mem' 'localmem' 'coalesced'
     'atomic' 'transfer' 'wgsize' 'transfer1' 'wgsize1' 'transfer.1'
     'wgsize.1' 'oracle' 'runtime_cpu' 'runtime_gpu' 'src' 'seq']

print(data.columns.values)
    data.rename(columns={ 'index-given':'u7777777','wgsize':'wwww'}, inplace = True) #替换列表列标签
    print(data.columns.values)

['index-given' 'dataset' 'comp' 'rational' 'mem' 'localmem' 'coalesced'
     'atomic' 'transfer' 'wgsize' 'transfer1' 'wgsize1' 'transfer.1'
     'wgsize.1' 'oracle' 'runtime_cpu' 'runtime_gpu' 'src' 'seq']
    ['u7777777' 'dataset' 'comp' 'rational' 'mem' 'localmem' 'coalesced'
     'atomic' 'transfer' 'wwww' 'transfer1' 'wgsize1' 'transfer.1' 'wgsize.1'
     'oracle' 'runtime_cpu' 'runtime_gpu' 'src' 'seq']

insert1 = data["u7777777"].values
    data.insert(5,'typeSmallSmall',insert1) #在第五个空子(0,1,2,3,4,5)插入列名字为typeSmallSmall的数据insert1
    print(data.columns.values)

['u7777777' 'dataset' 'comp' 'rational' 'mem' 'typeSmallSmall' 'localmem'
     'coalesced' 'atomic' 'transfer' 'wwww' 'transfer1' 'wgsize1' 'transfer.1'
     'wgsize.1' 'oracle' 'runtime_cpu' 'runtime_gpu' 'src' 'seq']