Merge branch 'develop' into feature/modify_priceticker

.github/workflows/autotrade.yml

@@ -21,7 +21,7 @@ jobs:
         with:
           node-version: 18
       - name: Install dependencies
-        run: npm ci
+        run: npm i
       - name: Build nestjs app
         run: npm run build
       - name: Run tests

.gitignore

@@ -34,4 +34,6 @@ lerna-debug.log*
 !.vscode/launch.json
 !.vscode/extensions.json
 request.http
+
+# package-lock.json
 package-lock.json

README.md

@@ -176,7 +176,7 @@ Parameters
 
 ```typescript
 
-PUT /tradebot?id=tradebotId
+GET /tradebot?id=tradebotId
 
 ```
 

package.json

@@ -27,6 +27,7 @@
     "@nestjs/mapped-types": "*",
     "@nestjs/platform-express": "^10.0.0",
     "@nestjs/schedule": "^4.0.1",
+    "@tensorflow/tfjs-node": "^4.17.0",
     "arima": "^0.2.5",
     "axios": "^1.5.1",
     "bignumber.js": "^9.1.2",

src/price_ticker/price_ticker.service.spec.ts

@@ -2,6 +2,7 @@ import { Test, TestingModule } from '@nestjs/testing';
 import PriceTickerService from './price_ticker.service';
 import { HttpModule } from '@nestjs/axios';
 import { QuotationDto } from './dto/quotation.dto';
+import * as fs from 'fs';
 
 describe('PriceTickerService', () => {
   let service: PriceTickerService;
@@ -36,4 +37,21 @@ describe('PriceTickerService', () => {
       .mockImplementation(async () => array);
     expect(await service.getCandlesticks()).toStrictEqual([1, 2, 3, 4, 5]);
   });
+
+  it('should return an array of candlestick', async () => {
+    let etharr = [];
+    for (let i = 0; i < 11; i++) {
+      const begin = Date.now() - 90000000 * (i + 1);
+      const end = Date.now() - 90000000 * i;
+      const r = await service.getCandlesticksV2('ETH-USDT', '5m', begin, end);
+      const tempArr = r.data.candlesticks.candlesticks.map(
+        (item) => item.y.close,
+      );
+      etharr = tempArr.concat(etharr);
+    }
+    const etharrJson = JSON.stringify(etharr);
+    // console.log(etharr);
+    // use fs to write the etharr to a file
+    fs.writeFileSync('src/strategies/etharr.txt', etharrJson);
+  });
 });

src/strategies/ETH-USD.csv

@@ -364,5 +364,4 @@ Date,Open,High,Low,Close,Adj Close,Volume
 2024-03-10,3915.590576,3968.723633,3800.564453,3881.193115,3881.193115,15783924355
 2024-03-11,3881.237793,4087.050049,3745.125244,4066.445068,4066.445068,28806262507
 2024-03-12,4066.690430,4092.284180,3831.889893,3980.273193,3980.273193,26917010932
-2024-03-13,3980.265137,4083.007324,3936.627197,4006.457031,4006.457031,22028114691
-2024-03-14,null,null,null,null,null,null
+2024-03-13,3980.265137,4083.007324,3936.627197,4006.457031,4006.457031,22028114691

src/strategies/strategies.controller.ts

@@ -1,7 +1,13 @@
-import { Controller } from '@nestjs/common';
+import { Controller, Get } from '@nestjs/common';
 import { StrategiesService } from './strategies.service';
 
 @Controller('strategies')
 export class StrategiesController {
   constructor(private readonly strategiesService: StrategiesService) {}
+
+  @Get()
+  async trainDqn() {
+    await this.strategiesService.trainDqn();
+    return 'Training DQN finished';
+  }
 }

src/strategies/strategies.service.spec.ts

@@ -116,58 +116,22 @@ describe('StrategiesService', () => {
     // Info: (20240320 - Jacky) this is aim to sum the profit of all trades
     const profitArray = tradeArray.map((trade) => trade.profit);
     const sum = profitArray.reduce((total, profit) => total + profit, 0);
-    expect(sum).toBe(-1069.9531315100012);
+    expect(sum).toBe(-1060.9868554650009);
   });
 
   it('should backtest use api', async () => {
     console.log = () => {};
-    const ETHPriceArray = [
-      3250.92, 3227.38, 3264, 3252.43, 3277.55, 3264.01, 3280.58, 3260.4,
-      3258.96, 3275.94, 3292.73, 3281.7, 3273.62, 3259.7, 3222.2, 3228.48,
-      3225.21, 3229.59, 3228, 3230.22, 3245.07, 3244.62, 3227.1, 3213.34,
-      3205.78, 3218.81, 3212.59, 3216.6, 3243.82, 3254.6, 3260.18, 3276.6,
-      3273.45, 3276.47, 3270, 3289.41, 3278.61, 3257.55, 3272.34, 3264.35,
-      3278.03, 3292.56, 3303.6, 3295.36, 3305.09, 3295.91, 3282.62, 3279.36,
-      3279.6, 3273.79, 3296.25, 3306.6, 3285.4, 3295.09, 3279.6, 3272.93,
-      3277.75, 3292.59, 3260, 3242.66, 3241.78, 3237.8, 3222.6, 3234.3, 3238.34,
-      3252.28, 3267.21, 3264.54, 3258.5, 3259.2, 3250.44, 3248.98, 3235.4,
-      3245.99, 3257.86, 3251.08, 3277.35, 3295.41, 3312.28, 3323.59, 3308.63,
-      3302.98, 3306.66, 3317.32, 3303.59, 3299.27, 3296.26, 3299, 3302.28,
-      3315.59, 3324.42, 3323.63, 3332.26, 3327.73, 3338.18, 3332.01, 3341.54,
-      3335.91, 3342.99, 3350.09, 3331.74, 3322.51, 3317.57, 3312.87, 3312.96,
-      3314.18, 3307.64, 3295.61, 3287, 3304.74, 3301.41, 3306.23, 3311.47,
-      3312.52, 3329.5, 3314.01, 3312.61, 3316.38, 3306.79, 3329.9, 3332.61,
-      3342.5, 3356, 3344.52, 3334.21, 3342.99, 3339.01, 3332.32, 3327.49,
-      3333.23, 3312.47, 3297.72, 3297.61, 3287.31, 3272.79, 3276.98, 3293.86,
-      3295.18, 3278.81, 3291.58, 3279.67, 3279.21, 3283.9, 3272.82, 3275.05,
-      3282.55, 3262.38, 3273.59, 3273.32, 3280.56, 3284.08, 3298.75, 3294.55,
-      3291.02, 3285.03, 3278.96, 3252.62, 3246.18, 3233.95, 3226.14, 3243.15,
-      3229.47, 3211.81, 3216.67, 3201.42, 3211.38, 3184.59, 3177.5, 3170.81,
-      3174.33, 3155.81, 3159.2, 3173.39, 3162.93, 3162.81, 3175.59, 3163.63,
-      3158.4, 3158.05, 3173.82, 3168.86, 3193.39, 3201.67, 3206.98, 3199.53,
-      3213.39, 3215.19, 3213.22, 3220.43, 3235.22, 3220.58, 3212.89, 3211.78,
-      3188.78, 3168.61, 3188.41, 3174.71, 3175.2, 3164.39, 3179.45, 3187.09,
-      3175.39, 3187.27, 3195.52, 3203.25, 3214.48, 3210.22, 3218.77, 3216.85,
-      3221, 3221.37, 3225.99, 3216.23, 3224, 3230.16, 3227.6, 3224.2, 3240,
-      3251.13, 3263.32, 3253.84, 3248.38, 3232.99, 3210.99, 3197.31, 3160.85,
-      3174.81, 3193.03, 3191.8, 3187.52, 3162.31, 3160.3, 3121.62, 3130,
-      3119.86, 3133, 3128.61, 3128.84, 3125.61, 3114.78, 3102.2, 3104.79,
-      3069.41, 3074, 3087.97, 3096.29, 3098.59, 3092.15, 3095.56, 3110.15,
-      3113.2, 3119.41, 3128.17, 3124.95, 3136.15, 3142.86, 3140, 3137.91,
-      3144.45, 3143.04, 3146.88, 3140.86, 3148.23, 3154.01, 3160.57, 3165.67,
-      3174.26, 3184.2, 3184.82, 3181.76, 3210.49, 3205.48, 3218.07, 3217.91,
-      3210.38, 3221.28, 3220.03, 3219.03, 3220.01, 3224.01, 3228.11, 3212.89,
-      3211.79, 3224.31, 3216.85, 3207.8, 3224.44, 3220.49, 3212.7, 3219.26,
-    ];
+    const ethArrFile = fs.readFileSync('src/strategies/etharr.txt', 'utf8');
+    const etharr = JSON.parse(ethArrFile);
     const { tradeArray } = await strategiesService.backTesting(
       'autoArima',
       'autoArima',
       'autoArima',
-      ETHPriceArray,
+      etharr,
     );
     // Info: (20240320 - Jacky) this is aim to sum the profit of all trades
     const profitArray = tradeArray.map((trade) => trade.profit);
     const sum = profitArray.reduce((total, profit) => total + profit, 0);
-    expect(sum).toBe(-673.9419000000021);
+    expect(sum).toBe(-7023.318874999997);
   });
 });

src/strategies/strategies.service.ts

@@ -1,4 +1,8 @@
 import { Injectable } from '@nestjs/common';
+import { Environment } from './strategy/rl-dqn/environment';
+import { TradeAgent } from './strategy/rl-dqn/tradeAgent';
+import { train } from './strategy/rl-dqn/train';
+import * as fs from 'fs';
 
 @Injectable()
 export class StrategiesService {
@@ -93,4 +97,12 @@ export class StrategiesService {
     }
     return { tradeArray };
   }
+
+  async trainDqn() {
+    const ethArrFile = fs.readFileSync('src/strategies/etharr.txt', 'utf8');
+    const etharr = JSON.parse(ethArrFile);
+    const env = new Environment(etharr);
+    const tradeAgent = new TradeAgent(env);
+    await train(tradeAgent);
+  }
 }

src/strategies/strategy/autoArima.ts

@@ -66,7 +66,7 @@ export function stopLoss(data: {
   const AbsSpreadFee = Math.abs(data.spreadFee);
   const openPrice = data.openPrice;
   const currentPrice = data.currentPrice;
-  const stopLoss = AbsSpreadFee * 0.7;
+  const stopLoss = AbsSpreadFee * 1;
   const holdingStatus = data.holdingStatus;
   if (holdingStatus === 'BUY') {
     // Info: (20240328 - Jacky) Current Buy price is lower than the open Buy price

src/strategies/strategy/linear.ts

@@ -7,7 +7,7 @@ export function suggestion(data: {
   let suggestion = 'WAIT';
   const priceArray = data.priceArray;
   const currentPrice = data.currentPrice;
-  const lastHourPrice = priceArray[priceArray.length - 11];
+  const lastHourPrice = priceArray[priceArray.length - 35];
   if (currentPrice > lastHourPrice) {
     suggestion = 'BUY';
   }

src/strategies/strategy/rl-dqn/dqn.ts

@@ -0,0 +1,62 @@
+/**
+ * @license
+ * Copyright 2019 Google LLC. All Rights Reserved.
+ * Licensed under the Apache License, Version 2.0 (the "License");
+ * you may not use this file except in compliance with the License.
+ * You may obtain a copy of the License at
+ *
+ * http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ * =============================================================================
+ */
+
+import * as tf from '@tensorflow/tfjs-node';
+
+export function createDeepQNetwork(numActions) {
+  const model = tf.sequential();
+  model.add(
+    tf.layers.dense({ units: 32, activation: 'relu', inputShape: [104] }),
+  );
+  model.add(tf.layers.dense({ units: 64, activation: 'relu' }));
+  model.add(tf.layers.dense({ units: 128, activation: 'relu' }));
+  model.add(tf.layers.dropout({ rate: 0.25 }));
+  // model.add(tf.layers.dense({ units: numActions }));
+  model.add(tf.layers.dense({ units: numActions, activation: 'sigmoid' }));
+  return model;
+}
+
+/**
+ * Copy the weights from a source deep-Q network to another.
+ *
+ * @param {tf.LayersModel} destNetwork The destination network of weight
+ *   copying.
+ * @param {tf.LayersModel} srcNetwork The source network for weight copying.
+ */
+export function copyWeights(destNetwork, srcNetwork) {
+  // https://github.com/tensorflow/tfjs/issues/1807:
+  // Weight orders are inconsistent when the trainable attribute doesn't
+  // match between two `LayersModel`s. The following is a workaround.
+  // TODO(cais): Remove the workaround once the underlying issue is fixed.
+  let originalDestNetworkTrainable;
+  if (destNetwork.trainable !== srcNetwork.trainable) {
+    originalDestNetworkTrainable = destNetwork.trainable;
+    destNetwork.trainable = srcNetwork.trainable;
+  }
+
+  destNetwork.setWeights(srcNetwork.getWeights());
+
+  // Weight orders are inconsistent when the trainable attribute doesn't
+  // match between two `LayersModel`s. The following is a workaround.
+  // TODO(cais): Remove the workaround once the underlying issue is fixed.
+  // `originalDestNetworkTrainable` is null if and only if the `trainable`
+  // properties of the two LayersModel instances are the same to begin
+  // with, in which case nothing needs to be done below.
+  if (originalDestNetworkTrainable != null) {
+    destNetwork.trainable = originalDestNetworkTrainable;
+  }
+}

src/strategies/strategy/rl-dqn/environment.ts

@@ -0,0 +1,127 @@
+export class Environment {
+  current_step: number;
+  initial_balance: number;
+  balance: number;
+  action_space: any;
+  done: boolean;
+  spreadFee: number;
+  holdingStatus: number;
+  priceArray: number[];
+  openPrice: number;
+  currentPrice: number;
+  profitLog: number[];
+  actionLog: any[];
+  rewardLog: any[];
+
+  constructor(priceArray) {
+    this.current_step = 48;
+    this.initial_balance = 10000;
+    this.spreadFee = 30;
+    this.balance = this.initial_balance;
+    this.priceArray = priceArray;
+    this.action_space = [0, 1, 2, 3]; // WAIT, BUY, SELL, CLOSE
+    this.done = false;
+    this.holdingStatus = 0;
+    this.profitLog = [];
+    this.actionLog = [];
+    this.rewardLog = [];
+  }
+
+  reset() {
+    this.current_step = 48;
+    this.initial_balance = 10000;
+    // this.spreadFee = 30;
+    this.balance = this.initial_balance;
+    this.done = false;
+    this.holdingStatus = 0;
+    this.profitLog = [];
+    return this;
+  }
+
+  // return {
+  //   currentPrice: currentPrice,
+  //   openPrice: this.openPrice,
+  //   spreadFee: spreadFee,
+  //   profit: profit,
+  //   holdingStatus: 'WAIT',
+  // };
+  step(action: any) {
+    let profit = 0;
+    let reward = 0;
+    this.currentPrice = this.priceArray[this.current_step];
+    this.spreadFee = 0.005 * this.currentPrice;
+    if (action === 3) {
+      if (this.holdingStatus === 0) {
+        reward = -40;
+      }
+      if (this.holdingStatus === 1) {
+        profit = this.currentPrice - this.spreadFee - this.openPrice;
+        reward = profit;
+        reward += 10;
+        this.holdingStatus = 0;
+      }
+      if (this.holdingStatus === 2) {
+        profit = this.openPrice - (this.currentPrice + this.spreadFee);
+        reward = profit;
+        reward += 10;
+        this.holdingStatus = 0;
+      }
+    }
+    if (action === 1) {
+      if (this.holdingStatus === 0) {
+        const currentPriceStr = JSON.stringify(this.currentPrice);
+        this.openPrice = JSON.parse(currentPriceStr);
+        this.holdingStatus = 1;
+        reward = 25;
+        if (this.openPrice < this.priceArray[this.current_step + 10]) {
+          reward += 24;
+        } else {
+          reward += -14;
+        }
+      } else {
+        reward = -35;
+      }
+    }
+    if (action === 2) {
+      if (this.holdingStatus === 0) {
+        const currentPriceStr = JSON.stringify(this.currentPrice);
+        this.openPrice = JSON.parse(currentPriceStr);
+        this.holdingStatus = 2;
+        reward = 25;
+        if (this.openPrice > this.priceArray[this.current_step + 10]) {
+          reward += 20;
+        } else {
+          reward += -20;
+        }
+      } else {
+        reward = -35;
+      }
+    }
+    this.current_step += 1;
+    if (this.current_step >= this.priceArray.length) {
+      this.done = true;
+    }
+    if (action === 0) {
+      if (this.actionLog.slice(-24).every((a) => a === 0)) {
+        reward = -30;
+      }
+    }
+    return {
+      previousPrice: this.priceArray.slice(-100),
+      currentPrice: this.currentPrice,
+      openPrice: this.openPrice,
+      spreadFee: this.spreadFee,
+      holdingStatus: this.holdingStatus,
+      reward: reward,
+    };
+  }
+  getState() {
+    return [
+      this.priceArray.slice(-100),
+      this.currentPrice,
+      this.openPrice,
+      this.spreadFee,
+      this.holdingStatus,
+    ];
+  }
+}

src/strategies/strategy/rl-dqn/models/dqn/model.json

@@ -0,0 +1 @@
+{"modelTopology":{"class_name":"Sequential","config":{"name":"sequential_1","layers":[{"class_name":"Dense","config":{"units":32,"activation":"relu","use_bias":true,"kernel_initializer":{"class_name":"VarianceScaling","config":{"scale":1,"mode":"fan_avg","distribution":"normal","seed":null}},"bias_initializer":{"class_name":"Zeros","config":{}},"kernel_regularizer":null,"bias_regularizer":null,"activity_regularizer":null,"kernel_constraint":null,"bias_constraint":null,"name":"dense_Dense1","trainable":true,"batch_input_shape":[null,104],"dtype":"float32"}},{"class_name":"Dense","config":{"units":64,"activation":"relu","use_bias":true,"kernel_initializer":{"class_name":"VarianceScaling","config":{"scale":1,"mode":"fan_avg","distribution":"normal","seed":null}},"bias_initializer":{"class_name":"Zeros","config":{}},"kernel_regularizer":null,"bias_regularizer":null,"activity_regularizer":null,"kernel_constraint":null,"bias_constraint":null,"name":"dense_Dense2","trainable":true}},{"class_name":"Dense","config":{"units":128,"activation":"relu","use_bias":true,"kernel_initializer":{"class_name":"VarianceScaling","config":{"scale":1,"mode":"fan_avg","distribution":"normal","seed":null}},"bias_initializer":{"class_name":"Zeros","config":{}},"kernel_regularizer":null,"bias_regularizer":null,"activity_regularizer":null,"kernel_constraint":null,"bias_constraint":null,"name":"dense_Dense3","trainable":true}},{"class_name":"Dropout","config":{"rate":0.25,"noise_shape":null,"seed":null,"name":"dropout_Dropout1","trainable":true}},{"class_name":"Dense","config":{"units":4,"activation":"sigmoid","use_bias":true,"kernel_initializer":{"class_name":"VarianceScaling","config":{"scale":1,"mode":"fan_avg","distribution":"normal","seed":null}},"bias_initializer":{"class_name":"Zeros","config":{}},"kernel_regularizer":null,"bias_regularizer":null,"activity_regularizer":null,"kernel_constraint":null,"bias_constraint":null,"name":"dense_Dense4","trainable":true}}]},"keras_version":"tfjs-layers 4.17.0","backend":"tensor_flow.js"},"weightsManifest":[{"paths":["weights.bin"],"weights":[{"name":"dense_Dense1/kernel","shape":[104,32],"dtype":"float32"},{"name":"dense_Dense1/bias","shape":[32],"dtype":"float32"},{"name":"dense_Dense2/kernel","shape":[32,64],"dtype":"float32"},{"name":"dense_Dense2/bias","shape":[64],"dtype":"float32"},{"name":"dense_Dense3/kernel","shape":[64,128],"dtype":"float32"},{"name":"dense_Dense3/bias","shape":[128],"dtype":"float32"},{"name":"dense_Dense4/kernel","shape":[128,4],"dtype":"float32"},{"name":"dense_Dense4/bias","shape":[4],"dtype":"float32"}]}],"format":"layers-model","generatedBy":"TensorFlow.js tfjs-layers v4.17.0","convertedBy":null}