2018年更新:
ipython notebook 已更名为 jupyter notebook 。安装使用方法如下：
pandas 和 jupyter notebook 。jupyter notebook 命令就可以使用 jupyter notebook 。%matplotlib inline 命令。下面是以前的安装使用方法：
安装 pandas
sudo apt-get install build-essential python-dev
sudo apt-get install python-pandas python-tk
sudo apt-get install python-scipy python-matplotlib python-tables
sudo apt-get install python-numexpr python-xlrd python-statsmodels
sudo apt-get install python-openpyxl python-xlwt python-bs4
if use virtualenv before install matplotlib should install libpng-dev, libjpeg8-dev, libfreetype6-dev
安装 ipython-notebook
sudo pip install "ipython[notebook]"
sudo pip install pygments
ipython notebook运行 ipython-notebook 。如果使用matplotlib内嵌进网页中,那么需要运行:ipython notebook --matplotlib inline 。
In [59]:
import numpy as np
import pandas as pd
In [2]:
%matplotlib inline
In [3]:
# 查看 pandas 的版本
pd.__version__
Out[3]:
In [4]:
s = pd.Series([1,3,5,np.nan,6,8]);s
Out[4]:
In [5]:
dates = pd.date_range('20180101', periods=6);dates
Out[5]:
In [6]:
df = pd.DataFrame(np.random.randn(6,4), index=dates, columns=list('ABCD'));df
Out[6]:
In [7]:
# 可以使用字典来创建 DataFrame 。
# 如果字典的 Value 是单一值，那么会自动扩展。
# 如果字典的 Value 是列表或者 Series ，那么长度要保持一致。
# 如果字典中只有一个值有 Index ，那么会使用这个 Index 作为整个 DataFrame 的 Index 。
# 如果字典有多个 Index ，那么必须保持一致，否则会报错。
df2 = pd.DataFrame(
{ 'A' : 1.,
'B' : pd.Timestamp('20130102'),
'C' : pd.Series(1,index=list(range(2,6)),dtype='float32'),
'D' : np.array([3] * 4,dtype='int32'),
'E' : pd.Categorical(["test","train","test","train"]),
'F' : 'foo' }
)
df2
Out[7]:
In [8]:
# 对象类型
df2.dtypes
Out[8]:
In [9]:
# 查看头部数据
df.head()
Out[9]:
In [10]:
# 查看尾部数据
df.tail(2)
#head 和 tail 接受一个整数参数，缺省值为 5 。
Out[10]:
In [11]:
df.index
Out[11]:
In [12]:
df.columns
Out[12]:
In [13]:
df.values
Out[13]:
In [14]:
df.describe()
Out[14]:
In [15]:
df.T
Out[15]:
In [16]:
# 读入 CSV 格式数据
df_movies = pd.read_csv('datas/movies.csv', sep='\t', encoding='utf-8')
df_movies.head()
Out[16]:
In [17]:
df_movies = pd.read_csv('datas/movies.csv', sep='\t', encoding='utf-8',thousands=',',escapechar='$')
df_movies.head()
Out[17]:
pandas.read_excel(io, sheet_name=0, header=0, skiprows=None, skip_footer=0,
index_col=None, names=None, usecols=None, parse_dates=False,
date_parser=None, na_values=None, thousands=None,
convert_float=True, converters=None, dtype=None,
true_values=None, false_values=None, engine=None,
squeeze=False, **kwds)
Read an Excel table into a pandas DataFrame
io : string, path object (pathlib.Path or py._path.local.LocalPath), file-like object, pandas ExcelFile, or xlrd workbook. The string could be a URL. Valid URL schemes include http, ftp, s3, and file. For file URLs, a host is expected. For instance, a local file could be file://localhost/path/to/workbook.xlsx
sheet_name : string, int, mixed list of strings/ints, or None, default 0 Strings are used for sheet names, Integers are used in zero-indexed sheet positions. Lists of strings/integers are used to request multiple sheets. Specify None to get all sheets. str|int -> DataFrame is returned. list|None -> Dict of DataFrames is returned, with keys representing sheets. Available Cases • Defaults to 0 -> 1st sheet as a DataFrame • 1 -> 2nd sheet as a DataFrame • “Sheet1” -> 1st sheet as a DataFrame • [0,1,”Sheet5”] -> 1st, 2nd & 5th sheet as a dictionary of DataFrames • None -> All sheets as a dictionary of DataFrames
sheetname : string, int, mixed list of strings/ints, or None, default 0 Deprecated since version 0.21.0: Use sheet_name instead
header : int, list of ints, default 0 Row (0-indexed) to use for the column labels of the parsed DataFrame. If a list of integers is passed those row positions will be combined into a MultiIndex. Use None if there is no header.
skiprows : list-like Rows to skip at the beginning (0-indexed)
skip_footer : int, default 0 Rows at the end to skip (0-indexed)
index_col : int, list of ints, default None Column (0-indexed) to use as the row labels of the DataFrame. Pass None if there is no such column. If a list is passed, those columns will be combined into a MultiIndex. If a subset of data is selected with usecols, index_col is based on the subset.
names : array-like, default None List of column names to use. If file contains no header row, then you should explicitly pass header=None
converters : dict, default None Dict of functions for converting values in certain columns. Keys can either be integers or column labels, values are functions that take one input argument, the Excel cell content, and return the transformed content.
dtype : Type name or dict of column -> type, default None Data type for data or columns. E.g. {‘a’: np.float64, ‘b’: np.int32} Use object to preserve data as stored in Excel and not interpret dtype. If converters are specified, they will be applied INSTEAD of dtype conversion. New in version 0.20.0.
true_values : list, default None Values to consider as True New in version 0.19.0.
false_values : list, default None Values to consider as False New in version 0.19.0.
parse_cols : int or list, default None Deprecated since version 0.21.0: Pass in usecols instead.
usecols : int or list, default None • If None then parse all columns, • If int then indicates last column to be parsed • If list of ints then indicates list of column numbers to be parsed • If string then indicates comma separated list of Excel column letters and column ranges (e.g. “A:E” or “A,C,E:F”). Ranges are inclusive of both sides.
squeeze : boolean, default False If the parsed data only contains one column then return a Series
na_values : scalar, str, list-like, or dict, default None Additional strings to recognize as NA/NaN. If dict passed, specific per-column NA values. By default the following values are interpreted as NaN: ‘’, ‘#N/A’, ‘#N/A N/A’, ‘#NA’, ‘-1.#IND’, ‘-1.#QNAN’, ‘-NaN’, ‘-nan’, ‘1.#IND’, ‘1.#QNAN’, ‘N/A’, ‘NA’, ‘NULL’, ‘NaN’, ‘n/a’, ‘nan’, ‘null’.
thousands : str, default None Thousands separator for parsing string columns to numeric. Note that this parameter is only necessary for columns stored as TEXT in Excel, any numeric columns will automatically be parsed, regardless of display format.
keep_default_na : bool, default True If na_values are specified and keep_default_na is False the default NaN values are over-ridden, otherwise they’re appended to.
verbose : boolean, default False Indicate number of NA values placed in non-numeric columns
engine: string, default None If io is not a buffer or path, this must be set to identify io. Acceptable values are None or xlrd
convert_float : boolean, default True convert integral floats to int (i.e., 1.0 –> 1). If False, all numeric data will be read in as floats: Excel stores all numbers as floats internally
DataFrame from the passed in Excel file. See notes in sheet_name argument for more information on when a Dict of Dataframes is returned.
In [18]:
df = df_movies.copy()
df.head(3)
Out[18]:
In [19]:
#显示开头的数据，缺省显示 5 条
df.head()
Out[19]:
In [20]:
#显示开头的数据，指定显示 3 条
df.head(3)
Out[20]:
In [21]:
#显示末尾的数据，缺省显示 5 条
df.tail()
Out[21]:
In [22]:
#显示末尾的数据，缺省显示 2 条
df.tail(2)
Out[22]:
In [23]:
#只显示指定的行和列
df.iloc[[1,3,5],[0,1,2,3]]
Out[23]:
In [24]:
df.loc[[1,3,5],['Date', 'Gross']]
Out[24]:
In [25]:
df = df_movies.copy()
# 单元格赋值
# 单个单元格赋值
df.iloc[0, 6] = u'土豆之歌'
df.loc[df.index[1], u'Gross']= 999
df.head(3)
Out[25]:
In [26]:
# 多单个单元格赋值
df.loc[df.index[0:2], u'Gross'] = [100, 200]
df.head(3)
Out[26]:
In [27]:
df = df_movies.copy()
#用一个列表来显式地指定，列表长度必须与列数一致
# 示例 1
df.columns = [u'Row', u'Date', u'WeekDay', u'Day', u'Top10Gross', u'No1Moive', u'Gross']
df.head()
Out[27]:
In [28]:
# 示例 2 ：大写转小写
df.columns = [c.lower() for c in df.columns]
df.head()
Out[28]:
In [29]:
# 示例 1 ：小写转大写
df = df.rename(columns=lambda x: x.upper())
df.tail(3)
Out[29]:
In [30]:
# 示例 2 ：改变特定的列头
df = df.rename(columns={'DATE': u'日期', 'GROSS': u'票房'})
df.head()
Out[30]:
In [31]:
df.columns.to_series().groupby(df.dtypes).groups
Out[31]:
In [32]:
# 打印列类型(清晰打印中文)
types = df.columns.to_series().groupby(df.dtypes).groups
for key, value in types.items():
print(key,':\t', ','.join(value))
In [33]:
df = df_movies.copy()
# 方式一：在末尾添加
df['memo'] = pd.Series('', index=df.index)
df.head(3)
Out[33]:
In [34]:
# 方式二：在中间插入
df = df_movies.copy()
df.insert(loc=1, column=u'year', value=u'2015')
df.head(3)
Out[34]:
In [35]:
# 根据现有值生成一个新的列
df = df_movies.copy()
df.insert(loc = 5 , column=u'OtherGross', value=df[u'Top 10 Gross'] - df[u'Gross'])
df.head(3)
Out[35]:
In [36]:
# 根据现有值生成多个新的列
df = df_movies.copy()
def process_date_col(text):
#根据日期生成月份和日两个新的列
if pd.isnull(text):
month = day = np.nan
else:
month, day = text.split('.')
return pd.Series([month, day])
df[[u'month', u'day']] = df.Date.apply(process_date_col)
df.head()
Out[36]:
In [37]:
df = df_movies.copy()
#根据一列的值改变另一列
df[u'#1 Movie'] = df[u'#1 Movie'].apply(lambda x: x[::-1])
df.head(3)
Out[37]:
In [38]:
# 同时改变多个列的值
cols = [u'Gross', u'Top 10 Gross']
df[cols] = df[cols].applymap(lambda x: x/10000)
df.head(3)
Out[38]:
In [39]:
df = df_movies.copy()
# 添加一个空行
df = df.append(pd.Series(
[np.nan]*len(df.columns), # Fill cells with NaNs
index=df.columns),
ignore_index=True)
df.tail(3)
Out[39]:
In [40]:
# 计数有空值的行
nans = df.shape[0] - df.dropna().shape[0]
print(u'一共有 %d 行出现空值' % nans)
# 填充空值为`无`
df.fillna(value=u'无', inplace=True)
df.tail()
Out[40]:
In [41]:
df = df_movies.copy()
# 添加一个空行
df = df.append(pd.Series(
[np.nan]*len(df.columns), # Fill cells with NaNs
index=df.columns),
ignore_index=True)
# 根据某一列排序（由低到高）
df.sort_values(u'Gross', ascending=True, inplace=True)
df.head()
Out[41]:
In [42]:
# 排序后重新编制索引
df.index = range(1,len(df.index)+1)
df.head()
Out[42]:
In [43]:
df = df_movies.copy()
# 根据列类型过滤
# 只选择字符串型的列
df.loc[:, (df.dtypes == np.dtype('O')).values].head()
Out[43]:
In [44]:
# 选择 artifact 为空值的行
df.iloc[0, 6] = np.nan
df.iloc[3, 6] = np.nan
df[df[u'Gross'].isnull()].head()
Out[44]:
In [45]:
# 选择'Gross'为非空值的行
df[df[u'Gross'].notnull()].head()
Out[45]:
In [46]:
# 根据条件过滤
df[ (df[u'Day'] == u'Sat') | (df[u'Day#'] <= 32) ]
Out[46]:
In [47]:
df[ (df[u'Day'] == u'Sat') & (df[u'Day#'] <= 32) ]
Out[47]:
In [48]:
# 横向合并
df_a = df.filter(regex='D', axis=1);
print(df_a.head())
print('============================')
df_b = df.filter(regex='ss', axis=1);
print(df_b.head())
print('============================')
df_c = pd.concat([df_a, df_b],axis=1)
print(df_c.head())
In [49]:
ts = pd.Series(np.random.randn(1000), index=pd.date_range('1/1/2000', periods=1000))
ts = ts.cumsum()
ts.plot()
Out[49]:
In [50]:
df = df_movies.copy()
df.head()
Out[50]:
In [51]:
df.plot(x='Date', y=['Top 10 Gross', 'Gross'])
Out[51]:
In [52]:
df_1 = df_movies.copy()
df_2 = pd.DataFrame({u'#1 Movie':[u'American Sniper',
u'SpongeBob',
u'Fifty Shades of Grey'],
u'chs':[u'美国阻击手',
u'海绵宝宝',
u'五十度灰']})
df_1.head()
Out[52]:
In [53]:
df_2.head()
Out[53]:
In [54]:
pd.merge(df_1, df_2, on=u'#1 Movie').head()
Out[54]:
In [55]:
# 导出周六的数据，格式为 CSV
# df[ (df['Day'] == 'Sat') ].to_csv('test_tmp.csv', mode='w', encoding='utf-8', index=False)
#在前面的文件中追加周日的数据
# df[ (df['Day'] == 'Sun') ].to_csv('test_output.csv', mode='a', header=False, encoding='utf-8', index=False)
In [56]:
# 输出为 dict 格式
# DataFrame.to_dict可以接受 ‘dict’, ‘list’, ‘series’, ‘split’, ‘records’, ‘index’
df = pd.DataFrame({'AAA' : [4,5,6,7], 'ABB' : [10,20,30,40],'CCC' : [100,50,-30,-50]}); df
Out[56]:
In [57]:
df.to_dict('records')
Out[57]:
In [58]:
# 过滤包含指定字符的列
df_A = df.filter(regex='A', axis=1); df_A
Out[58]: