numpy操作方法

╰+攻爆jí腚メ 2022-02-19 11:45 240阅读 0赞

np.unique(array, return_inverse=True)

np_utils.to_categorical(coded_id)

np.random

随机数：

<table> 
 <tbody> 
  <tr> 
   <td style="width:196px;"> <p><a href="http://docs.scipy.org/doc/numpy/reference/generated/numpy.random.rand.html#numpy.random.rand" rel="nofollow">rand</a>(d0,&nbsp;d1,&nbsp;...,&nbsp;dn)</p> </td> 
   <td style="width:653px;"> <p>随机值；给定形状，生成[0,1]之间均匀分布的随机样本</p> <p>np.random.rand(2，3)&nbsp; &nbsp;[[0.42085 &nbsp;0.74822 0.79452][0.08833 0.85959 0.25843 ]]</p> <p>np.random.rand(3)&nbsp; &nbsp; &nbsp; &nbsp; &nbsp;[0.50288776 0.33155867 0.84370448]</p> </td> 
  </tr> 
  <tr> 
   <td style="width:196px;"> <p><a href="http://docs.scipy.org/doc/numpy/reference/generated/numpy.random.randn.html#numpy.random.randn" rel="nofollow">randn</a>(d0,&nbsp;d1,&nbsp;...,&nbsp;dn)</p> </td> 
   <td style="width:653px;"> <p>同rand,&nbsp;生成符合标准正态分布N(0,1)&nbsp;的随机样本；同<a href="https://docs.scipy.org/doc/numpy/reference/generated/numpy.random.standard_normal.html#numpy.random.standard_normal" rel="nofollow"><code>standard_normal</code></a>([size])</p> </td> 
  </tr> 
  <tr> 
   <td style="width:196px;">randint(low,high=None,size=None,dtype)&nbsp;</td> 
   <td style="width:653px;"> <p>生成在区间<strong>[low,high)</strong>上离散均匀分布的整数值;若<strong>high=None</strong>，则取值区间变为<strong>[0,low)</strong>&nbsp;</p> <p>np.random.randint(5, size=(2, 4))&nbsp; &nbsp;&nbsp;<strong>[[1 0 2 3]&nbsp;[4 3 2 4]]</strong></p> <p>np.random.randint(2, size=5)&nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp;<strong>[1 0 0 0 1]</strong></p> <p><span style="color:#000000;font-family:Consolas, Inconsolata, Courier, monospace;">np.random.randint(2,5,size=5) </span><span style="color:#000000;">[2 2 3 4 3]</span></p> </td> 
  </tr> 
  <tr> 
   <td style="width:196px;"> <p>random_integers(low,high=None,size=None)&nbsp;</p> </td> 
   <td style="width:653px;"> <p>生成在闭区间 [low, high]离散均匀分布的整数值。若<strong>high=None</strong>，则取值区间变为<strong>[1,low]</strong>&nbsp;</p> <p>np.random.random_integers(1,5,size=(2,3))&nbsp; &nbsp;[[4 2 4]&nbsp;[1 1 5]]</p> </td> 
  </tr> 
  <tr> 
   <td style="width:196px;"> <p><a href="http://docs.scipy.org/doc/numpy/reference/generated/numpy.random.random_sample.html#numpy.random.random_sample" rel="nofollow">random_sample</a>([size])</p> </td> 
   <td style="width:653px;"> <p>给定形状，返回随机的浮点数，在半开区间 [0.0, 1.0)。</p> </td> 
  </tr> 
  <tr> 
   <td style="width:196px;"> <p><a href="http://docs.scipy.org/doc/numpy/reference/generated/numpy.random.random.html#numpy.random.random" rel="nofollow">random</a>([size])</p> </td> 
   <td style="width:653px;"> <p>返回随机的浮点数，在半开区间 [0.0, 1.0)。（官网例子与random_sample完全一样）</p> </td> 
  </tr> 
  <tr> 
   <td style="width:196px;"> <p><a href="http://docs.scipy.org/doc/numpy/reference/generated/numpy.random.ranf.html#numpy.random.ranf" rel="nofollow">ranf</a>([size])</p> </td> 
   <td style="width:653px;"> <p>返回随机的浮点数，在半开区间 [0.0, 1.0)。（官网例子与random_sample完全一样）</p> </td> 
  </tr> 
  <tr> 
   <td style="width:196px;"> <p><a href="http://docs.scipy.org/doc/numpy/reference/generated/numpy.random.sample.html#numpy.random.sample" rel="nofollow">sample</a>([size])</p> </td> 
   <td style="width:653px;"> <p>返回随机的浮点数，在半开区间 [0.0, 1.0)。（官网例子与random_sample完全一样）</p> </td> 
  </tr> 
  <tr> 
   <td style="width:196px;">choice(a,size=None,replace=True,p=None)&nbsp;</td> 
   <td style="width:653px;"> <p>若a为数组，则从a中选取元素；</p> <p>若a为单个int类型数，则选取range(a)中的数&nbsp;<br> replace是bool类型，为True，则选取的元素会出现重复；反之不会出现重复&nbsp;<br> p为数组，里面存放选到每个数的可能性，即概率&nbsp;</p> </td> 
  </tr> 
  <tr> 
   <td style="width:196px;"> <p><a href="http://docs.scipy.org/doc/numpy/reference/generated/numpy.random.bytes.html#numpy.random.bytes" rel="nofollow">bytes</a>(length)</p> </td> 
   <td style="width:653px;"> <p>返回随机字节。&gt;&gt;&gt; np.random.bytes(10)</p> </td> 
  </tr> 
 </tbody> 
</table>

排列

<table> 
 <tbody> 
  <tr> 
   <td> <p><a href="http://docs.scipy.org/doc/numpy/reference/generated/numpy.random.shuffle.html#numpy.random.shuffle" rel="nofollow">shuffle</a>(x)</p> </td> 
   <td> <p>现场修改序列，改变自身内容。（类似洗牌，打乱顺序） np.random.shuffle(arr)</p> </td> 
  </tr> 
  <tr> 
   <td> <p><a href="http://docs.scipy.org/doc/numpy/reference/generated/numpy.random.permutation.html#numpy.random.permutation" rel="nofollow">permutation</a>(x)</p> </td> 
   <td> <p>返回一个随机排列&nbsp; &nbsp;np.random.permutation(10)&nbsp; &nbsp;<span style="color:#000000;">np.random.permutation(array)</span></p> </td> 
  </tr> 
 </tbody> 
</table>

随机数生成器：

<table> 
 <tbody> 
  <tr> 
   <td> <p><a href="http://docs.scipy.org/doc/numpy/reference/generated/numpy.random.RandomState.html#numpy.random.RandomState" rel="nofollow">RandomState</a></p> </td> 
   <td>Container for the Mersenne Twister pseudo-random number generator.</td> 
  </tr> 
  <tr> 
   <td> <p><a href="http://docs.scipy.org/doc/numpy/reference/generated/numpy.random.seed.html#numpy.random.seed" rel="nofollow">seed</a>([seed])</p> </td> 
   <td>Seed the generator.</td> 
  </tr> 
  <tr> 
   <td> <p><a href="http://docs.scipy.org/doc/numpy/reference/generated/numpy.random.get_state.html#numpy.random.get_state" rel="nofollow">get_state</a>()</p> </td> 
   <td>Return a tuple representing the internal state of the generator.</td> 
  </tr> 
  <tr> 
   <td> <p><a href="http://docs.scipy.org/doc/numpy/reference/generated/numpy.random.set_state.html#numpy.random.set_state" rel="nofollow">set_state</a>(state)</p> </td> 
   <td>Set the internal state of the generator from a tuple.</td> 
  </tr> 
 </tbody> 
</table>

分布：

<table> 
 <tbody> 
  <tr> 
   <td> <p><a href="http://docs.scipy.org/doc/numpy/reference/generated/numpy.random.beta.html#numpy.random.beta" rel="nofollow">beta</a>(a,&nbsp;b[,&nbsp;size])</p> </td> 
   <td>贝塔分布样本，在&nbsp;[0,&nbsp;1]内。</td> 
  </tr> 
  <tr> 
   <td> <p><a href="http://docs.scipy.org/doc/numpy/reference/generated/numpy.random.binomial.html#numpy.random.binomial" rel="nofollow">binomial</a>(n,&nbsp;p[,&nbsp;size])</p> </td> 
   <td>二项分布的样本。</td> 
  </tr> 
  <tr> 
   <td> <p><a href="http://docs.scipy.org/doc/numpy/reference/generated/numpy.random.chisquare.html#numpy.random.chisquare" rel="nofollow">chisquare</a>(df[,&nbsp;size])</p> </td> 
   <td>卡方分布样本。</td> 
  </tr> 
  <tr> 
   <td> <p><a href="http://docs.scipy.org/doc/numpy/reference/generated/numpy.random.dirichlet.html#numpy.random.dirichlet" rel="nofollow">dirichlet</a>(alpha[,&nbsp;size])</p> </td> 
   <td>狄利克雷分布样本。</td> 
  </tr> 
  <tr> 
   <td> <p><a href="http://docs.scipy.org/doc/numpy/reference/generated/numpy.random.exponential.html#numpy.random.exponential" rel="nofollow">exponential</a>([scale,&nbsp;size])</p> </td> 
   <td>指数分布</td> 
  </tr> 
  <tr> 
   <td> <p><a href="http://docs.scipy.org/doc/numpy/reference/generated/numpy.random.f.html#numpy.random.f" rel="nofollow">f</a>(dfnum,&nbsp;dfden[,&nbsp;size])</p> </td> 
   <td>F分布样本。</td> 
  </tr> 
  <tr> 
   <td> <p><a href="http://docs.scipy.org/doc/numpy/reference/generated/numpy.random.gamma.html#numpy.random.gamma" rel="nofollow">gamma</a>(shape[,&nbsp;scale,&nbsp;size])</p> </td> 
   <td>伽马分布</td> 
  </tr> 
  <tr> 
   <td> <p><a href="http://docs.scipy.org/doc/numpy/reference/generated/numpy.random.geometric.html#numpy.random.geometric" rel="nofollow">geometric</a>(p[,&nbsp;size])</p> </td> 
   <td>几何分布</td> 
  </tr> 
  <tr> 
   <td> <p><a href="http://docs.scipy.org/doc/numpy/reference/generated/numpy.random.gumbel.html#numpy.random.gumbel" rel="nofollow">gumbel</a>([loc,&nbsp;scale,&nbsp;size])</p> </td> 
   <td>耿贝尔分布。</td> 
  </tr> 
  <tr> 
   <td> <p><a href="http://docs.scipy.org/doc/numpy/reference/generated/numpy.random.hypergeometric.html#numpy.random.hypergeometric" rel="nofollow">hypergeometric</a>(ngood,&nbsp;nbad,&nbsp;nsample[,&nbsp;size])</p> </td> 
   <td>超几何分布样本。</td> 
  </tr> 
  <tr> 
   <td> <p><a href="http://docs.scipy.org/doc/numpy/reference/generated/numpy.random.laplace.html#numpy.random.laplace" rel="nofollow">laplace</a>([loc,&nbsp;scale,&nbsp;size])</p> </td> 
   <td>拉普拉斯或双指数分布样本</td> 
  </tr> 
  <tr> 
   <td> <p><a href="http://docs.scipy.org/doc/numpy/reference/generated/numpy.random.logistic.html#numpy.random.logistic" rel="nofollow">logistic</a>([loc,&nbsp;scale,&nbsp;size])</p> </td> 
   <td>Logistic分布样本</td> 
  </tr> 
  <tr> 
   <td> <p><a href="http://docs.scipy.org/doc/numpy/reference/generated/numpy.random.lognormal.html#numpy.random.lognormal" rel="nofollow">lognormal</a>([mean,&nbsp;sigma,&nbsp;size])</p> </td> 
   <td>对数正态分布</td> 
  </tr> 
  <tr> 
   <td> <p><a href="http://docs.scipy.org/doc/numpy/reference/generated/numpy.random.logseries.html#numpy.random.logseries" rel="nofollow">logseries</a>(p[,&nbsp;size])</p> </td> 
   <td>对数级数分布。</td> 
  </tr> 
  <tr> 
   <td> <p><a href="http://docs.scipy.org/doc/numpy/reference/generated/numpy.random.multinomial.html#numpy.random.multinomial" rel="nofollow">multinomial</a>(n,&nbsp;pvals[,&nbsp;size])</p> </td> 
   <td>多项分布</td> 
  </tr> 
  <tr> 
   <td> <p><a href="http://docs.scipy.org/doc/numpy/reference/generated/numpy.random.multivariate_normal.html#numpy.random.multivariate_normal" rel="nofollow">multivariate_normal</a>(mean,&nbsp;cov[,&nbsp;size])</p> </td> 
   <td> <p>多元正态分布。</p> <pre><code>&gt;&gt;&gt; mean = [0,0]
&gt;&gt;&gt; cov = [[1,0],[0,100]] 
&gt;&gt;&gt; import matplotlib.pyplot as plt
&gt;&gt;&gt; x, y = np.random.multivariate_normal(mean, cov, 5000).T
&gt;&gt;&gt; plt.plot(x, y, ‘x‘); plt.axis(‘equal‘); plt.show()
</code></pre> </td> 
  </tr> 
  <tr> 
   <td> <p><a href="http://docs.scipy.org/doc/numpy/reference/generated/numpy.random.negative_binomial.html#numpy.random.negative_binomial" rel="nofollow">negative_binomial</a>(n,&nbsp;p[,&nbsp;size])</p> </td> 
   <td>负二项分布</td> 
  </tr> 
  <tr> 
   <td> <p><a href="http://docs.scipy.org/doc/numpy/reference/generated/numpy.random.noncentral_chisquare.html#numpy.random.noncentral_chisquare" rel="nofollow">noncentral_chisquare</a>(df,&nbsp;nonc[,&nbsp;size])</p> </td> 
   <td>非中心卡方分布</td> 
  </tr> 
  <tr> 
   <td> <p><a href="http://docs.scipy.org/doc/numpy/reference/generated/numpy.random.noncentral_f.html#numpy.random.noncentral_f" rel="nofollow">noncentral_f</a>(dfnum,&nbsp;dfden,&nbsp;nonc[,&nbsp;size])</p> </td> 
   <td>非中心F分布</td> 
  </tr> 
  <tr> 
   <td> <p><a href="http://docs.scipy.org/doc/numpy/reference/generated/numpy.random.normal.html#numpy.random.normal" rel="nofollow">normal</a>([loc,&nbsp;scale,&nbsp;size])</p> </td> 
   <td> <p>正态(高斯)分布</p> <pre><code>&gt;&gt;&gt; mu, sigma = 0, 0.1 # mean and standard deviation
&gt;&gt;&gt; s = np.random.normal(mu, sigma, 1000)
&gt;&gt;&gt; abs(mu - np.mean(s)) &lt; 0.01
&gt;&gt;&gt; abs(sigma - np.std(s, ddof=1)) &lt; 0.01
</code></pre> </td> 
  </tr> 
  <tr> 
   <td> <p><a href="http://docs.scipy.org/doc/numpy/reference/generated/numpy.random.pareto.html#numpy.random.pareto" rel="nofollow">pareto</a>(a[,&nbsp;size])</p> </td> 
   <td>帕累托（Lomax）分布</td> 
  </tr> 
  <tr> 
   <td> <p><a href="http://docs.scipy.org/doc/numpy/reference/generated/numpy.random.poisson.html#numpy.random.poisson" rel="nofollow">poisson</a>([lam,&nbsp;size])</p> </td> 
   <td>泊松分布</td> 
  </tr> 
  <tr> 
   <td> <p><a href="http://docs.scipy.org/doc/numpy/reference/generated/numpy.random.power.html#numpy.random.power" rel="nofollow">power</a>(a[,&nbsp;size])</p> </td> 
   <td>Draws samples in [0, 1] from a power distribution with positive exponent a - 1.</td> 
  </tr> 
  <tr> 
   <td> <p><a href="http://docs.scipy.org/doc/numpy/reference/generated/numpy.random.rayleigh.html#numpy.random.rayleigh" rel="nofollow">rayleigh</a>([scale,&nbsp;size])</p> </td> 
   <td>Rayleigh&nbsp;分布</td> 
  </tr> 
  <tr> 
   <td> <p><a href="http://docs.scipy.org/doc/numpy/reference/generated/numpy.random.standard_cauchy.html#numpy.random.standard_cauchy" rel="nofollow">standard_cauchy</a>([size])</p> </td> 
   <td>标准柯西分布</td> 
  </tr> 
  <tr> 
   <td> <p><a href="http://docs.scipy.org/doc/numpy/reference/generated/numpy.random.standard_exponential.html#numpy.random.standard_exponential" rel="nofollow">standard_exponential</a>([size])</p> </td> 
   <td>标准的指数分布</td> 
  </tr> 
  <tr> 
   <td> <p><a href="http://docs.scipy.org/doc/numpy/reference/generated/numpy.random.standard_gamma.html#numpy.random.standard_gamma" rel="nofollow">standard_gamma</a>(shape[,&nbsp;size])</p> </td> 
   <td>标准伽马分布</td> 
  </tr> 
  <tr> 
   <td> <p><a href="http://docs.scipy.org/doc/numpy/reference/generated/numpy.random.standard_normal.html#numpy.random.standard_normal" rel="nofollow">standard_normal</a>([size])</p> </td> 
   <td>标准正态分布&nbsp;(mean=0, stdev=1).</td> 
  </tr> 
  <tr> 
   <td> <p><a href="http://docs.scipy.org/doc/numpy/reference/generated/numpy.random.standard_t.html#numpy.random.standard_t" rel="nofollow">standard_t</a>(df[,&nbsp;size])</p> </td> 
   <td>Standard Student’s t distribution with df degrees of freedom.</td> 
  </tr> 
  <tr> 
   <td> <p><a href="http://docs.scipy.org/doc/numpy/reference/generated/numpy.random.triangular.html#numpy.random.triangular" rel="nofollow">triangular</a>(left,&nbsp;mode,&nbsp;right[,&nbsp;size])</p> </td> 
   <td>三角形分布</td> 
  </tr> 
  <tr> 
   <td> <p><a href="http://docs.scipy.org/doc/numpy/reference/generated/numpy.random.uniform.html#numpy.random.uniform" rel="nofollow">uniform</a>([low,&nbsp;high,&nbsp;size])</p> </td> 
   <td>均匀分布</td> 
  </tr> 
  <tr> 
   <td> <p><a href="http://docs.scipy.org/doc/numpy/reference/generated/numpy.random.vonmises.html#numpy.random.vonmises" rel="nofollow">vonmises</a>(mu,&nbsp;kappa[,&nbsp;size])</p> </td> 
   <td>von Mises分布</td> 
  </tr> 
  <tr> 
   <td> <p><a href="http://docs.scipy.org/doc/numpy/reference/generated/numpy.random.wald.html#numpy.random.wald" rel="nofollow">wald</a>(mean,&nbsp;scale[,&nbsp;size])</p> </td> 
   <td>瓦尔德（逆高斯）分布</td> 
  </tr> 
  <tr> 
   <td> <p><a href="http://docs.scipy.org/doc/numpy/reference/generated/numpy.random.weibull.html#numpy.random.weibull" rel="nofollow">weibull</a>(a[,&nbsp;size])</p> </td> 
   <td>Weibull&nbsp;分布</td> 
  </tr> 
  <tr> 
   <td> <p><a href="http://docs.scipy.org/doc/numpy/reference/generated/numpy.random.zipf.html#numpy.random.zipf" rel="nofollow">zipf</a>(a[,&nbsp;size])</p> </td> 
   <td>齐普夫分布</td> 
  </tr> 
 </tbody> 
</table>

数组属性：

<table> 
 <thead> 
  <tr> 
   <th>属性</th> 
   <th>说明</th> 
  </tr> 
 </thead> 
 <tbody> 
  <tr> 
   <td>ndarray.ndim</td> 
   <td>秩，即轴的数量或维度的数量</td> 
  </tr> 
  <tr> 
   <td>ndarray.shape</td> 
   <td>数组的维度，对于矩阵，n 行 m 列</td> 
  </tr> 
  <tr> 
   <td>ndarray.size</td> 
   <td>数组元素的总个数，相当于 .shape 中 n*m 的值</td> 
  </tr> 
  <tr> 
   <td>ndarray.dtype</td> 
   <td>ndarray 对象的元素类型</td> 
  </tr> 
  <tr> 
   <td>ndarray.itemsize</td> 
   <td>ndarray 对象中每个元素的大小，以字节为单位</td> 
  </tr> 
  <tr> 
   <td>ndarray.flags</td> 
   <td>ndarray 对象的内存信息</td> 
  </tr> 
  <tr> 
   <td>ndarray.real</td> 
   <td>ndarray元素的实部</td> 
  </tr> 
  <tr> 
   <td>ndarray.imag</td> 
   <td>ndarray 元素的虚部</td> 
  </tr> 
  <tr> 
   <td>ndarray.data</td> 
   <td>包含实际数组元素的缓冲区，由于一般通过数组的索引获取元素，所以通常不需要使用这个属性。</td> 
  </tr> 
 </tbody> 
</table>

创建数组：

1. numpy.empty(shape, dtype = float, order = 'C')    数组元素为随机值

2. numpy.zeros(shape, dtype = float, order = 'C')   0数组

3. numpy.asarray(a, dtype = None, order = None)    a  list/tuple/list_tuple

4. numpy.frombuffer(buffer, dtype = float, count = -1, offset = 0)  动态数组，流输入 buffer='Hello World'

5. numpy.fromiter(iterable, dtype, count=-1)   迭代对象  iterable=iter([1,2,3,4,5])

6. numpy.arange(start, stop, step, dtype)  范围创建

7. np.linspace(start, stop, num=50, endpoint=True, retstep=False, dtype=None)  等差数列

8. np.logspace(start, stop, num=50, endpoint=True, base=10.0, dtype=None)  等比数列  base 参数意思是取对数的时候 log 的下标  start=base ** start  stop=base ** stop
    
    9. for x in np.nditer(a.T, order='C')  C行序优先  F列序优先

修改数组形状：

<table> 
 <thead> 
  <tr> 
   <th>函数</th> 
   <th>描述</th> 
  </tr> 
 </thead> 
 <tbody> 
  <tr> 
   <td><code>reshape</code></td> 
   <td>不改变数据的条件下修改形状&nbsp;&nbsp;np.arange(8).reshape(4,2)&nbsp; &nbsp;共享内存，同时修改</td> 
  </tr> 
  <tr> 
   <td><code>flat</code></td> 
   <td>数组元素迭代器&nbsp;&nbsp;for element in a.flat</td> 
  </tr> 
  <tr> 
   <td><code>flatten</code></td> 
   <td>返回一份数组拷贝，对拷贝所做的修改不会影响原始数组&nbsp; &nbsp; &nbsp;ndarray.flatten(order='C')</td> 
  </tr> 
  <tr> 
   <td><code>ravel</code></td> 
   <td> <p>返回展开数组,返回view(),会影响源数组&nbsp; &nbsp;numpy.ravel(a, order='C')</p> </td> 
  </tr> 
 </tbody> 
</table>

翻转数组：

<table> 
 <tbody> 
  <tr> 
   <td><code>transpose</code></td> 
   <td>对换数组的维度&nbsp; &nbsp; numpy.transpose(arr, axes)</td> 
  </tr> 
  <tr> 
   <td><code>ndarray.T</code></td> 
   <td>和&nbsp;<code>self.transpose()</code>&nbsp;相同&nbsp; &nbsp; &nbsp;a.T</td> 
  </tr> 
  <tr> 
   <td><code>rollaxis</code></td> 
   <td>向后滚动指定的轴&nbsp; &nbsp;numpy.rollaxis(arr, axis, start)</td> 
  </tr> 
  <tr> 
   <td><code>swapaxes</code></td> 
   <td>对换数组的两个轴&nbsp; numpy.swapaxes(arr, axis1, axis2)</td> 
  </tr> 
 </tbody> 
</table>

修改数组维度：

<table> 
 <thead> 
  <tr> 
   <th style="width:202px;">维度</th> 
   <th style="width:647px;">描述</th> 
  </tr> 
 </thead> 
 <tbody> 
  <tr> 
   <td style="width:202px;"><code>broadcast</code></td> 
   <td style="width:647px;">产生模仿广播的对象</td> 
  </tr> 
  <tr> 
   <td style="width:202px;"><code>broadcast_to</code></td> 
   <td style="width:647px;">将数组广播到新形状&nbsp; &nbsp;np.broadcast_to(a,(4,4))</td> 
  </tr> 
  <tr> 
   <td style="width:202px;"><code>expand_dims</code></td> 
   <td style="width:647px;">扩展数组的形状&nbsp; &nbsp;numpy.expand_dims(arr, axis)&nbsp; &nbsp; arr:输入数组&nbsp; axis:插入轴位置</td> 
  </tr> 
  <tr> 
   <td style="width:202px;"><code>squeeze</code></td> 
   <td style="width:647px;">从数组的形状中删除一维条目&nbsp;&nbsp;</td> 
  </tr> 
 </tbody> 
</table>

连接数组：

<table> 
 <thead> 
  <tr> 
   <th style="width:206px;">函数</th> 
   <th style="width:643px;">描述</th> 
  </tr> 
 </thead> 
 <tbody> 
  <tr> 
   <td style="width:206px;"><code>concatenate</code></td> 
   <td style="width:643px;">连接沿现有轴的数组序列&nbsp; &nbsp;numpy.concatenate((a1, a2, ...), axis)</td> 
  </tr> 
  <tr> 
   <td style="width:206px;"><code>stack</code></td> 
   <td style="width:643px;">沿着新的轴加入一系列数组。 numpy.stack(arrays, axis)</td> 
  </tr> 
  <tr> 
   <td style="width:206px;"><code>hstack</code></td> 
   <td style="width:643px;">水平堆叠序列中的数组（列方向）</td> 
  </tr> 
  <tr> 
   <td style="width:206px;"><code>vstack</code></td> 
   <td style="width:643px;">竖直堆叠序列中的数组（行方向）</td> 
  </tr> 
 </tbody> 
</table>

分割数组：

<table> 
 <thead> 
  <tr> 
   <th style="width:152px;">函数</th> 
   <th style="width:697px;">数组及操作</th> 
  </tr> 
 </thead> 
 <tbody> 
  <tr> 
   <td style="width:152px;"><code>split</code></td> 
   <td style="width:697px;">将一个数组分割为多个子数组&nbsp; numpy.split(ary, indices_or_sections, axis)&nbsp;&nbsp;np.split(a,3)&nbsp;&nbsp;np.split(a,[4,7])</td> 
  </tr> 
  <tr> 
   <td style="width:152px;"><code>hsplit</code></td> 
   <td style="width:697px;">将一个数组水平分割为多个子数组（按列）</td> 
  </tr> 
  <tr> 
   <td style="width:152px;"><code>vsplit</code></td> 
   <td style="width:697px;">将一个数组垂直分割为多个子数组（按行）</td> 
  </tr> 
 </tbody> 
</table>

数组元素的添加与删除

<table> 
 <thead> 
  <tr> 
   <th style="width:112px;">函数</th> 
   <th>元素及描述</th> 
  </tr> 
 </thead> 
 <tbody> 
  <tr> 
   <td style="width:112px;"><code>resize</code></td> 
   <td>返回指定形状的新数组&nbsp; &nbsp;如果新数组大小大于原始大小，则包含原始数组中的元素的副本。</td> 
  </tr> 
  <tr> 
   <td style="width:112px;"><code>append</code></td> 
   <td>将值添加到数组末尾&nbsp; numpy.append(arr, values, axis=None)&nbsp;axis无定义时，横向加成返回一维数组</td> 
  </tr> 
  <tr> 
   <td style="width:112px;"><code>insert</code></td> 
   <td>沿指定轴将值插入到指定下标之前 numpy.insert(arr, obj, values, axis)&nbsp; obj：索引</td> 
  </tr> 
  <tr> 
   <td style="width:112px;"><code>delete</code></td> 
   <td>删掉某个轴的子数组，并返回删除后的新数组&nbsp; &nbsp;Numpy.delete(arr, obj, axis)</td> 
  </tr> 
  <tr> 
   <td style="width:112px;"><code>unique</code></td> 
   <td>查找数组内的唯一元素&nbsp; &nbsp; u,indices = <span style="color:#000000;">numpy.unique(arr, return_index, return_inverse, return_counts)</span></td> 
  </tr> 
 </tbody> 
</table>

字符串函数

<table> 
 <thead> 
  <tr> 
   <th style="width:158px;">函数</th> 
   <th style="width:691px;">描述</th> 
  </tr> 
 </thead> 
 <tbody> 
  <tr> 
   <td style="width:158px;"><code>add()</code></td> 
   <td style="width:691px;">对两个数组的逐个字符串元素进行连接&nbsp;&nbsp;numpy.char.add()&nbsp;&nbsp;np.char.add(['hello'],[' xyz'])</td> 
  </tr> 
  <tr> 
   <td style="width:158px;">multiply()</td> 
   <td style="width:691px;">返回按元素多重连接后的字符串&nbsp;&nbsp;numpy.char.multiply()&nbsp; &nbsp;&nbsp;np.char.multiply('Runoob ',3)</td> 
  </tr> 
  <tr> 
   <td style="width:158px;"><code>center()</code></td> 
   <td style="width:691px;">居中字符串，并使用指定字符在左侧和右侧进行填充。&nbsp; np.char.center('Runoob', 20,fillchar = '*')</td> 
  </tr> 
  <tr> 
   <td style="width:158px;"><code>capitalize()</code></td> 
   <td style="width:691px;">将字符串第一个字母转换为大写&nbsp;&nbsp;np.char.capitalize('runoob')</td> 
  </tr> 
  <tr> 
   <td style="width:158px;"><code>title()</code></td> 
   <td style="width:691px;">将字符串的每个单词的第一个字母转换为大写&nbsp;&nbsp;np.char.title('i like runoob')</td> 
  </tr> 
  <tr> 
   <td style="width:158px;"><code>lower()</code></td> 
   <td style="width:691px;">数组元素转换为小写&nbsp;&nbsp;np.char.lower('RUNOOB')</td> 
  </tr> 
  <tr> 
   <td style="width:158px;"><code>upper()</code></td> 
   <td style="width:691px;">数组元素转换为大写&nbsp;&nbsp;np.char.upper(['runoob','google'])</td> 
  </tr> 
  <tr> 
   <td style="width:158px;"><code>split()</code></td> 
   <td style="width:691px;">指定分隔符对字符串进行分割，并返回数组列表&nbsp;&nbsp;np.char.split ('www.runoob.com', sep = '.')</td> 
  </tr> 
  <tr> 
   <td style="width:158px;"><code>splitlines()</code></td> 
   <td style="width:691px;">返回元素中的行列表，以换行符分割&nbsp;&nbsp;np.char.splitlines('i\nlike runoob?')</td> 
  </tr> 
  <tr> 
   <td style="width:158px;"><code>strip()</code></td> 
   <td style="width:691px;">移除元素开头或者结尾处的特定字符&nbsp;&nbsp;np.char.strip(['arunooba','admin','java'],'a')</td> 
  </tr> 
  <tr> 
   <td style="width:158px;"><code>join()</code></td> 
   <td style="width:691px;">通过指定分隔符来连接数组中的元素&nbsp;&nbsp;np.char.join([':','-'],['runoob','google'])&nbsp; &nbsp;np.char.join(':','runoob')</td> 
  </tr> 
  <tr> 
   <td style="width:158px;"><code>replace()</code></td> 
   <td style="width:691px;">使用新字符串替换字符串中的所有子字符串&nbsp;&nbsp;np.char.replace ('i like runoob', 'oo', 'cc')</td> 
  </tr> 
  <tr> 
   <td style="width:158px;"><code>decode()</code></td> 
   <td style="width:691px;">数组元素依次调用<code>str.decode&nbsp;&nbsp;</code>np.char.encode('runoob', 'cp500')</td> 
  </tr> 
  <tr> 
   <td style="width:158px;"><code>encode()</code></td> 
   <td style="width:691px;">数组元素依次调用<code>str.encode&nbsp;&nbsp;</code>np.char.decode(a,'cp500')</td> 
  </tr> 
 </tbody> 
</table>

函数：

舍入函数：numpy.around(a,decimals)
    
      三角函数：sin()、cos()、tan()
    
      下舍整数： numpy.floor()
    
      上入整数： numpy.ceil() 
    
      倒数： numpy.reciprocal() 函数返回参数逐元素的倒数
    
      幂函数： numpy.power()   np.power(a,2) np.power(a,b) [10,100,1000][1,2,3]
    
      加减乘除：add()，subtract()，multiply() 和 divide() np.add(a,b)
    
      余数： numpy.mod()  numpy.remainder() 
    
      最大值最小值（axis=None,0,1）： numpy.amin() 和 numpy.amax()
    
      最大值最小值差： numpy.ptp()
    
      百分位数，表示小于这个值的观察值的百分比: numpy.percentile(a, q, axis)
    
      中位数： numpy.median() 
     
      均值： numpy.mean() 
     
      加权平均： np.average([1,2,3, 4],weights = [4,3,2,1], returned = True) returned 参数设为 true，则返回权重的和
    
      标准差： std = sqrt(mean((x - x.mean())**2)) ；  np.std([1,2,3,4])
    
      方差：np.var([1,2,3,4])

排序：

<table> 
 <thead> 
  <tr> 
   <th>种类</th> 
   <th>速度</th> 
   <th>最坏情况</th> 
   <th>工作空间</th> 
   <th>稳定性</th> 
  </tr> 
 </thead> 
 <tbody> 
  <tr> 
   <td><code>'quicksort'</code>（快速排序）</td> 
   <td>1</td> 
   <td><code>O(n^2)</code></td> 
   <td>0</td> 
   <td>否</td> 
  </tr> 
  <tr> 
   <td><code>'mergesort'</code>（归并排序）</td> 
   <td>2</td> 
   <td><code>O(n*log(n))</code></td> 
   <td>~n/2</td> 
   <td>是</td> 
  </tr> 
  <tr> 
   <td><code>'heapsort'</code>（堆排序）</td> 
   <td>3</td> 
   <td><code>O(n*log(n))</code></td> 
   <td>0</td> 
   <td>否</td> 
  </tr> 
 </tbody> 
</table>

numpy.sort(a, axis, kind, order)
    
      numpy.argsort() 函数返回的是数组值从小到大的索引值
    
      numpy.argmax() 和 numpy.argmin()函数分别沿给定轴返回最大和最小元素的索引
    
      numpy.nonzero() 函数返回输入数组中非零元素的索引。
    
      numpy.where() 函数返回输入数组中满足给定条件的元素的索引。
    
      numpy.lexsort() 用于对多个序列进行排序
    
      numpy.extract() 函数根据某个条件从数组中抽取元素，返回满条件的元素。

<table> 
 <tbody> 
  <tr> 
   <th>函数</th> 
   <th>描述</th> 
  </tr> 
  <tr> 
   <td>msort(a)</td> 
   <td>数组按第一个轴排序，返回排序后的数组副本。np.msort(a) 相等于 np.sort(a, axis=0)。</td> 
  </tr> 
  <tr> 
   <td>sort_complex(a)</td> 
   <td>对复数按照先实部后虚部的顺序进行排序。</td> 
  </tr> 
  <tr> 
   <td>partition(a, kth[, axis, kind, order])</td> 
   <td>指定一个数，对数组进行分区</td> 
  </tr> 
  <tr> 
   <td>argpartition(a, kth[, axis, kind, order])</td> 
   <td>可以通过关键字 kind 指定算法沿着指定轴对数组进行分区</td> 
  </tr> 
 </tbody> 
</table>

副本和视图：  
  
简单的赋值不会创建数组对象的副本。     b = a  
ndarray.view() 方会创建一个新的数组对象。  
使用切片创建视图修改数据会影响到原始数组  
ndarray.copy() 函数创建一个副本 。

矩阵：

numpy.matlib.empty(shape, dtype, order)
    numpy.matlib.zeros((2,2))
    numpy.matlib.ones((2,2)) 
    numpy.matlib.eye(n, M,k, dtype)  对角线元素为 1，其他位置为零
    numpy.matlib.identity(5, dtype = float) 函数返回给定大小的单位矩阵
    numpy.matlib.rand(3,3) 函数创建一个给定大小的矩阵，数据是随机填充的。

线性代数：

<table> 
 <thead> 
  <tr> 
   <th>函数</th> 
   <th>描述</th> 
  </tr> 
 </thead> 
 <tbody> 
  <tr> 
   <td><code>dot</code></td> 
   <td>两个数组的点积，即元素对应相乘。</td> 
  </tr> 
  <tr> 
   <td><code>vdot</code></td> 
   <td>两个向量的点积</td> 
  </tr> 
  <tr> 
   <td><code>inner</code></td> 
   <td>两个数组的内积</td> 
  </tr> 
  <tr> 
   <td><code>matmul</code></td> 
   <td>两个数组的矩阵积</td> 
  </tr> 
  <tr> 
   <td><code>determinant</code></td> 
   <td>数组的行列式</td> 
  </tr> 
  <tr> 
   <td><code>solve</code></td> 
   <td>求解线性矩阵方程</td> 
  </tr> 
  <tr> 
   <td><code>inv</code></td> 
   <td>计算矩阵的乘法逆矩阵</td> 
  </tr> 
 </tbody> 
</table>

matplotlib：
    加载下载的字体：# zhfont1 = matplotlib.font_manager.FontProperties(fname="SimHei.ttf") 
    系统字体：a=sorted([f.name for f in matplotlib.font_manager.fontManager.ttflist])
    
    plt.title("Matplotlib demo") 
    plt.xlabel("x axis caption") 
    plt.plot(x,y,"ob") 
    plt.show()
    
    plt.subplot(2, 1, 1) 
    plt.subplot(2, 1, 2)
    
    plt.bar(x, y, align = 'center') 
    plt.bar(x2, y2, color = 'g', align = 'center') 
    
    hist,bins = np.histogram(a,bins = [0,20,40,60,80,100])  数据的频率分布的图形表示。 水平尺寸相等的矩形对应于类间隔，称为 bin，变量 height 对应于频率。
    plt.hist(a, bins = [0,20,40,60,80,100])