暑假更新

Big_data_example/scala/scala0224/src/main/java/com/atguigu/scala/AAA.scala

@@ -0,0 +1,10 @@
+package com.atguigu.scala
+
+class AAA(bbb:BBB) {
+
+  def open(): Unit = {
+    bbb.open()
+    println("aaaOpen")
+  }
+
+}

Big_data_example/scala/scala0224/src/main/java/com/atguigu/scala/BBB.scala

@@ -0,0 +1,15 @@
+package com.atguigu.scala
+
+class BBB {
+
+  def this(a:Int,b:Int) {
+    this()
+    println(a)
+    println(b)
+  }
+
+  def open()  ={
+    println("open")
+  }
+
+}

Big_data_example/scala/scala0224/src/main/java/com/atguigu/scala/test111.scala

@@ -0,0 +1,9 @@
+package com.atguigu.scala
+
+object test111 {
+
+  def main(args: Array[String]): Unit = {
+    new AAA(new BBB(1,2)).open()
+  }
+
+}

Leecode/src/main/java/com/markilue/leecode/Test1.java

@@ -2,7 +2,11 @@ package com.markilue.leecode;
 
 import cn.hutool.json.JSONObject;
 import cn.hutool.json.JSONUtil;
-import com.cqu.phmapiclientsdk.client.PhmApiClient;
+
+
+import java.time.LocalDate;
+import java.time.LocalDateTime;
+import java.time.format.DateTimeFormatter;
 
 /**
  *@BelongsProject: Leecode
@@ -15,9 +19,9 @@ import com.cqu.phmapiclientsdk.client.PhmApiClient;
 public class Test1 {
 
     public static void main(String[] args) {
-        PhmApiClient phmApiClient = new PhmApiClient("875f0554b4e041b3ba4be50cb68eb94f", "438fdae618794937a5333cfb6856a601");
+        String dateString = "20230822";
-        JSONObject jsonObject = phmApiClient.invokeByURL("http://172.28.9.61:8090/api/gas-turbine/things/search/", "{\"page_index\": 1, \"page_size\": 10, \"thing_group_uuid\":\"c1b3eb711fbd4475b262fe84fc965ea1\"}", "POST");
+        DateTimeFormatter formatter = DateTimeFormatter.ofPattern("yyyyMMdd");
-        String s = JSONUtil.toJsonStr(jsonObject);
+        LocalDate dateTime = LocalDate.parse(dateString, formatter);
-        System.out.println(s);
+        System.out.println("Parsed LocalDateTime: " + dateTime);
     }
 }

Leecode/src/main/java/com/markilue/leecode/test/Fibonaqi.java

@@ -16,7 +16,7 @@ public class Fibonaqi {
     public static void main(String[] args) {
         int n = 5;
         System.out.println(fibonacci(1, 1, 10));
-        Arrays.copyOf()
+
     }
 
     public static int fibonacci(int first, int second, int n) {

TensorFlow_eaxmple/Model_train_test/RUL/otherIdea/lru/lru.py

@@ -0,0 +1,117 @@
+#! -*- coding: utf-8 -*-
+# 线性循环单元（Linear Recurrent Unit）
+# tensorflow 1.15 + bert4keras 0.11.4 测试通过
+
+from bert4keras.layers import *
+
+
+class LRU(Layer):
+    """线性循环单元
+    链接1：https://arxiv.org/abs/2303.06349
+    链接2：https://kexue.fm/archives/9554
+    """
+    def __init__(
+        self,
+        units,
+        activation='linear',
+        use_bias=True,
+        unroll=True,  # unroll可以加速训练，但是会增加显存消耗
+        kernel_initializer='glorot_uniform',
+        **kwargs
+    ):
+        super(LRU, self).__init__(**kwargs)
+        self.units = units
+        self.activation = activations.get(activation)
+        self.use_bias = use_bias
+        self.unroll = unroll
+        self.kernel_initializer = initializers.get(kernel_initializer)
+
+    @integerize_shape
+    def build(self, input_shape):
+        super(LRU, self).build(input_shape)
+        hidden_size = input_shape[-1]
+        self.i_dense = Dense(
+            units=self.units * 2,
+            use_bias=self.use_bias,
+            kernel_initializer=self.kernel_initializer
+        )
+        self.o_dense = Dense(
+            units=hidden_size,
+            use_bias=self.use_bias,
+            activation=self.activation,
+            kernel_initializer=self.kernel_initializer
+        )
+
+        def initializer(shape, dtype=None):
+            r_min, r_max = 0.9, 0.999
+            u1 = np.random.random(size=shape[1])
+            u2 = np.random.random(size=shape[1])
+            nu_log = np.log(
+                -0.5 * np.log(u1 * (r_max**2 - r_min**2) + r_min**2)
+            )
+            theta_log = np.log(u2 * np.pi * 2)
+            gamma_log = np.log(np.sqrt(1 - np.exp(-np.exp(nu_log))**2))
+            return np.array([nu_log, theta_log, gamma_log])
+
+        self.params_log = self.add_weight(
+            name='params_log', shape=(3, self.units), initializer=initializer
+        )
+
+    @recompute_grad
+    def call(self, inputs, mask=None):
+        u = self.i_dense(inputs)
+        params = K.exp(self.params_log)
+        nu, theta, gamma = params[0], params[1], params[2]
+
+        if self.unroll:
+            L_in = K.int_shape(u)[1]
+            assert L_in is not None, 'input_length can not be None while unroll=True'
+            log2_L = int(np.ceil(np.log2(L_in)))
+        else:
+            L_in = K.shape(u)[1]
+            log2_L = K.log(K.cast(L_in, K.floatx())) / K.log(2.)
+            log2_L = K.cast(tf.ceil(log2_L), 'int32')
+
+        u = tf.complex(u[..., ::2], u[..., 1::2])
+        u = tf.pad(u, [[0, 0], [0, 2**log2_L - K.shape(u)[1]], [0, 0]])
+        B, L, D = K.shape(u)[0], K.shape(u)[1], K.int_shape(u)[-1]
+
+        def lru(i, x):
+            l = 2**i
+            x = K.reshape(x, [B * L // l, l, D])
+            x1, x2 = x[:, :l // 2], x[:, l // 2:]
+
+            pos = K.arange(1, l // 2 + 1, dtype=K.floatx())
+            nus = tf.einsum('n,d->nd', pos, nu)
+            thetas = tf.einsum('n,d->nd', pos, theta)
+            lambs = K.exp(tf.complex(-nus, thetas))
+
+            x2 = x2 + lambs * x1[:, -1:]
+            x = K.concatenate([x1, x2], axis=1)
+            if (not self.unroll) and K.int_shape(u)[1] is not None:
+                x = K.reshape(x, [B, L, D])
+
+            return i + 1, x
+
+        if self.unroll:
+            x = u
+            for i in range(log2_L):
+                _, x = lru(i + 1, x)
+        else:
+            _, x = tf.while_loop(lambda i, x: i <= log2_L, lru, [1, u])
+
+        x = x[:, :L_in] * tf.complex(gamma, 0.)
+        x = K.concatenate([tf.real(x), tf.imag(x)], axis=-1)
+        return self.o_dense(x)
+
+    def get_config(self):
+        config = {
+            'units': self.units,
+            'activation': activations.serialize(self.activation),
+            'use_bias': self.use_bias,
+            'unroll': self.unroll,
+            'kernel_initializer':
+                initializers.serialize(self.kernel_initializer),
+        }
+        base_config = super(LRU, self).get_config()
+        return dict(list(base_config.items()) + list(config.items()))

TensorFlow_eaxmple/Model_train_test/RUL/otherIdea/lru/lru2.0.py

@@ -0,0 +1,113 @@
+import numpy as np
+import tensorflow as tf
+from tensorflow.keras.layers import Layer, Dense, activations, initializers
+from tensorflow.keras import backend as K
+
+class LRU(Layer):
+    """线性循环单元
+    链接1：https://arxiv.org/abs/2303.06349
+    链接2：https://kexue.fm/archives/9554
+    """
+    def __init__(
+        self,
+        units,
+        activation='linear',
+        use_bias=True,
+        unroll=True,  # unroll可以加速训练，但是会增加显存消耗
+        kernel_initializer='glorot_uniform',
+        **kwargs
+    ):
+        super(LRU, self).__init__(**kwargs)
+        self.units = units
+        self.activation = activations.get(activation)
+        self.use_bias = use_bias
+        self.unroll = unroll
+        self.kernel_initializer = initializers.get(kernel_initializer)
+
+    def build(self, input_shape):
+        super(LRU, self).build(input_shape)
+        hidden_size = input_shape[-1]
+        self.i_dense = Dense(
+            units=self.units * 2,
+            use_bias=self.use_bias,
+            kernel_initializer=self.kernel_initializer
+        )
+        self.o_dense = Dense(
+            units=hidden_size,
+            use_bias=self.use_bias,
+            activation=self.activation,
+            kernel_initializer=self.kernel_initializer
+        )
+
+        def initializer(shape, dtype=None):
+            r_min, r_max = 0.9, 0.999
+            u1 = np.random.random(size=shape[1])
+            u2 = np.random.random(size=shape[1])
+            nu_log = np.log(
+                -0.5 * np.log(u1 * (r_max**2 - r_min**2) + r_min**2)
+            )
+            theta_log = np.log(u2 * np.pi * 2)
+            gamma_log = np.log(np.sqrt(1 - np.exp(-np.exp(nu_log))**2))
+            return np.array([nu_log, theta_log, gamma_log])
+
+        self.params_log = self.add_weight(
+            name='params_log', shape=(3, self.units), initializer=initializer
+        )
+
+    @tf.function
+    def call(self, inputs, mask=None):
+        u = self.i_dense(inputs)
+        params = tf.exp(self.params_log)
+        nu, theta, gamma = params[0], params[1], params[2]
+
+        if self.unroll:
+            L_in = K.int_shape(u)[1]
+            assert L_in is not None, 'input_length can not be None while unroll=True'
+            log2_L = tf.cast(tf.math.ceil(tf.math.log(L_in) / tf.math.log(2.)), tf.int32)
+        else:
+            L_in = tf.shape(u)[1]
+            log2_L = tf.cast(tf.math.ceil(tf.math.log(tf.cast(L_in, dtype=tf.float32)) / tf.math.log(2.)), tf.int32)
+
+        u = tf.complex(u[..., ::2], u[..., 1::2])
+        u = tf.pad(u, [[0, 0], [0, 2**log2_L - tf.shape(u)[1]], [0, 0]])
+        B, L, D = tf.shape(u)[0], tf.shape(u)[1], tf.shape(u)[-1]
+
+        def lru(i, x):
+            l = 2**i
+            x = tf.reshape(x, [B * L // l, l, D])
+            x1, x2 = x[:, :l // 2], x[:, l // 2:]
+
+            pos = tf.range(1, l // 2 + 1, dtype=tf.float32)
+            nus = tf.einsum('n,d->nd', pos, nu)
+            thetas = tf.einsum('n,d->nd', pos, theta)
+            lambs = tf.exp(tf.complex(-nus, thetas))
+
+            x2 = x2 + lambs * x1[:, -1:]
+            x = tf.concat([x1, x2], axis=1)
+            if (not self.unroll) and tf.shape(u)[1] is not None:
+                x = tf.reshape(x, [B, L, D])
+
+            return i + 1, x
+
+        if self.unroll:
+            x = u
+            for i in range(log2_L):
+                _, x = lru(i + 1, x)
+        else:
+            _, x = tf.while_loop(lambda i, x: i <= log2_L, lru, [1, u])
+
+        x = x[:, :L_in] * tf.complex(gamma, 0.)
+        x = tf.concat([tf.math.real(x), tf.math.imag(x)], axis=-1)
+        return self.o_dense(x)
+
+    def get_config(self):
+        config = {
+            'units': self.units,
+            'activation': activations.serialize(self.activation),
+            'use_bias': self.use_bias,
+            'unroll': self.unroll,
+            'kernel_initializer':
+                initializers.serialize(self.kernel_initializer),
+        }
+        base_config = super(LRU, self).get_config()
+        return dict(list(base_config.items()) + list(config.items()))