python matplotlib 绘制二维数据中某些列到折线图,没有线的解决方法

其他 2021-01-27 12:12:00 阅读次数: 0

results.txt中的文本数据是这样的.

{
    
     "epoch":0, "imgindex":0,  "imgcount":63,  "lr":0.0002 ,  "loss":3.2448 ,  "loss_classifier":2.5456 ,  "loss_box_reg":0.0000 ,  "loss_mask":0.0000 ,  "loss_objectness":0.6903 ,  "loss_rpn_box_reg":0.0089 , "time":31.9483, "data":12.9320, "GPUMemory":2177 }
{
    
     "epoch":0, "imgindex":20, "imgcount":63,  "lr":0.0034 ,  "loss":3.6288 ,  "loss_classifier":2.2065 ,  "loss_box_reg":0.0035 ,  "loss_mask":0.7142 ,  "loss_objectness":0.6809 ,  "loss_rpn_box_reg":0.0142 , "time":1.7869, "data":0.1222, "GPUMemory":3140 }
{
    
     "epoch":0, "imgindex":40, "imgcount":63,  "lr":0.0066 ,  "loss":1.0768 ,  "loss_classifier":0.0840 ,  "loss_box_reg":0.0093 ,  "loss_mask":0.7100 ,  "loss_objectness":0.1529 ,  "loss_rpn_box_reg":0.0186 , "time":0.7657, "data":0.1278, "GPUMemory":3140 }
{
    
     "epoch":0, "imgindex":60, "imgcount":63,  "lr":0.0098 ,  "loss":0.8712 ,  "loss_classifier":0.0616 ,  "loss_box_reg":0.0064 ,  "loss_mask":0.6750 ,  "loss_objectness":0.0762 ,  "loss_rpn_box_reg":0.0127 , "time":0.7116, "data":0.0756, "GPUMemory":3156 }
{
    
     "epoch":0, "imgindex":62, "imgcount":63,  "lr":0.0100 ,  "loss":0.8712 ,  "loss_classifier":0.0616 ,  "loss_box_reg":0.0069 ,  "loss_mask":0.6750 ,  "loss_objectness":0.0704 ,  "loss_rpn_box_reg":0.0112 , "time":0.7035, "data":0.0756, "GPUMemory":3156 }
{
    
     "epoch":1, "imgindex":0,  "imgcount":63,  "lr":0.0100 ,  "loss":0.8573 ,  "loss_classifier":0.0519 ,  "loss_box_reg":0.0147 ,  "loss_mask":0.7056 ,  "loss_objectness":0.0710 ,  "loss_rpn_box_reg":0.0140 , "time":12.6762, "data":12.0246, "GPUMemory":3156 }
{
    
     "epoch":1, "imgindex":20, "imgcount":63,  "lr":0.0100 ,  "loss":0.8407 ,  "loss_classifier":0.0779 ,  "loss_box_reg":0.0193 ,  "loss_mask":0.6626 ,  "loss_objectness":0.0539 ,  "loss_rpn_box_reg":0.0151 , "time":0.7691, "data":0.1215, "GPUMemory":3156 }
{
    
     "epoch":1, "imgindex":40, "imgcount":63,  "lr":0.0100 ,  "loss":0.7553 ,  "loss_classifier":0.0395 ,  "loss_box_reg":0.0011 ,  "loss_mask":0.6084 ,  "loss_objectness":0.0866 ,  "loss_rpn_box_reg":0.0156 , "time":0.6852, "data":0.0516, "GPUMemory":3156 }
{
    
     "epoch":1, "imgindex":60, "imgcount":63,  "lr":0.0100 ,  "loss":0.8489 ,  "loss_classifier":0.0594 ,  "loss_box_reg":0.0133 ,  "loss_mask":0.6666 ,  "loss_objectness":0.0681 ,  "loss_rpn_box_reg":0.0146 , "time":0.8134, "data":0.1637, "GPUMemory":3156 }
{
    
     "epoch":1, "imgindex":62, "imgcount":63,  "lr":0.0100 ,  "loss":0.8576 ,  "loss_classifier":0.0601 ,  "loss_box_reg":0.0166 ,  "loss_mask":0.6531 ,  "loss_objectness":0.0699 ,  "loss_rpn_box_reg":0.0146 , "time":0.8066, "data":0.1636, "GPUMemory":3156 }
{
    
     "epoch":2, "imgindex":0,  "imgcount":63,  "lr":0.0100 ,  "loss":0.1494 ,  "loss_classifier":0.0470 ,  "loss_box_reg":0.0000 ,  "loss_mask":0.0000 ,  "loss_objectness":0.0822 ,  "loss_rpn_box_reg":0.0202 , "time":13.3777, "data":12.7521, "GPUMemory":3156 }
{
    
     "epoch":2, "imgindex":20, "imgcount":63,  "lr":0.0100 ,  "loss":0.8790 ,  "loss_classifier":0.0629 ,  "loss_box_reg":0.0055 ,  "loss_mask":0.6817 ,  "loss_objectness":0.0797 ,  "loss_rpn_box_reg":0.0121 , "time":0.8539, "data":0.2024, "GPUMemory":3156 }
{
    
     "epoch":2, "imgindex":40, "imgcount":63,  "lr":0.0100 ,  "loss":0.7611 ,  "loss_classifier":0.0461 ,  "loss_box_reg":0.0101 ,  "loss_mask":0.6072 ,  "loss_objectness":0.0531 ,  "loss_rpn_box_reg":0.0119 , "time":0.6516, "data":0.0061, "GPUMemory":3156 }
{
    
     "epoch":2, "imgindex":60, "imgcount":63,  "lr":0.0100 ,  "loss":0.8804 ,  "loss_classifier":0.0712 ,  "loss_box_reg":0.0410 ,  "loss_mask":0.6410 ,  "loss_objectness":0.0652 ,  "loss_rpn_box_reg":0.0173 , "time":0.6745, "data":0.0346, "GPUMemory":3156 }
{
    
     "epoch":2, "imgindex":62, "imgcount":63,  "lr":0.0100 ,  "loss":0.8526 ,  "loss_classifier":0.0663 ,  "loss_box_reg":0.0350 ,  "loss_mask":0.6372 ,  "loss_objectness":0.0652 ,  "loss_rpn_box_reg":0.0170 , "time":0.6737, "data":0.0334, "GPUMemory":3156 }
{
    
     "epoch":3, "imgindex":0,  "imgcount":63,  "lr":0.0100 ,  "loss":1.0212 ,  "loss_classifier":0.1881 ,  "loss_box_reg":0.0681 ,  "loss_mask":0.6798 ,  "loss_objectness":0.0648 ,  "loss_rpn_box_reg":0.0204 , "time":12.8021, "data":12.1545, "GPUMemory":3156 }
{
    
     "epoch":3, "imgindex":20, "imgcount":63,  "lr":0.0100 ,  "loss":0.8134 ,  "loss_classifier":0.0495 ,  "loss_box_reg":0.0202 ,  "loss_mask":0.6327 ,  "loss_objectness":0.0553 ,  "loss_rpn_box_reg":0.0143 , "time":0.7610, "data":0.1089, "GPUMemory":3156 }
{
    
     "epoch":3, "imgindex":40, "imgcount":63,  "lr":0.0100 ,  "loss":0.8319 ,  "loss_classifier":0.0645 ,  "loss_box_reg":0.0041 ,  "loss_mask":0.6497 ,  "loss_objectness":0.0677 ,  "loss_rpn_box_reg":0.0143 , "time":0.7960, "data":0.1426, "GPUMemory":3156 }
{
    
     "epoch":3, "imgindex":60, "imgcount":63,  "lr":0.0100 ,  "loss":0.8450 ,  "loss_classifier":0.0851 ,  "loss_box_reg":0.0353 ,  "loss_mask":0.6324 ,  "loss_objectness":0.0646 ,  "loss_rpn_box_reg":0.0112 , "time":0.6900, "data":0.0534, "GPUMemory":3156 }
{
    
     "epoch":3, "imgindex":62, "imgcount":63,  "lr":0.0100 ,  "loss":0.8455 ,  "loss_classifier":0.0882 ,  "loss_box_reg":0.0360 ,  "loss_mask":0.6351 ,  "loss_objectness":0.0534 ,  "loss_rpn_box_reg":0.0105 , "time":0.6847, "data":0.0535, "GPUMemory":3156 }
{
    
     "epoch":4, "imgindex":0,  "imgcount":63,  "lr":0.0100 ,  "loss":0.1181 ,  "loss_classifier":0.0425 ,  "loss_box_reg":0.0000 ,  "loss_mask":0.0000 ,  "loss_objectness":0.0597 ,  "loss_rpn_box_reg":0.0159 , "time":12.8091, "data":12.1755, "GPUMemory":3156 }
{
    
     "epoch":4, "imgindex":20, "imgcount":63,  "lr":0.0100 ,  "loss":0.8551 ,  "loss_classifier":0.0512 ,  "loss_box_reg":0.0236 ,  "loss_mask":0.5953 ,  "loss_objectness":0.0606 ,  "loss_rpn_box_reg":0.0136 , "time":0.8668, "data":0.2161, "GPUMemory":3156 }
{
    
     "epoch":4, "imgindex":40, "imgcount":63,  "lr":0.0100 ,  "loss":0.8365 ,  "loss_classifier":0.0867 ,  "loss_box_reg":0.0317 ,  "loss_mask":0.6267 ,  "loss_objectness":0.0467 ,  "loss_rpn_box_reg":0.0139 , "time":0.7096, "data":0.0670, "GPUMemory":3156 }
{
    
     "epoch":4, "imgindex":60, "imgcount":63,  "lr":0.0100 ,  "loss":0.8199 ,  "loss_classifier":0.1017 ,  "loss_box_reg":0.0467 ,  "loss_mask":0.5918 ,  "loss_objectness":0.0750 ,  "loss_rpn_box_reg":0.0159 , "time":0.7013, "data":0.0584, "GPUMemory":3156 }
{
    
     "epoch":4, "imgindex":62, "imgcount":63,  "lr":0.0100 ,  "loss":0.9040 ,  "loss_classifier":0.1018 ,  "loss_box_reg":0.0560 ,  "loss_mask":0.6074 ,  "loss_objectness":0.0750 ,  "loss_rpn_box_reg":0.0159 , "time":0.6953, "data":0.0575, "GPUMemory":3156 }
{
    
     "epoch":5, "imgindex":0, "imgcount":63,  "lr":0.0100 ,  "loss":0.0743 ,  "loss_classifier":0.0040 ,  "loss_box_reg":0.0000 ,  "loss_mask":0.0000 ,  "loss_objectness":0.0536 ,  "loss_rpn_box_reg":0.0168 , "time":12.8451, "data":12.2135, "GPUMemory":3156 }
{
    
     "epoch":5, "imgindex":20, "imgcount":63,  "lr":0.0100 ,  "loss":0.7978 ,  "loss_classifier":0.0431 ,  "loss_box_reg":0.0129 ,  "loss_mask":0.6468 ,  "loss_objectness":0.0534 ,  "loss_rpn_box_reg":0.0108 , "time":0.7597, "data":0.1117, "GPUMemory":3156 }
{
    
     "epoch":5, "imgindex":40, "imgcount":63,  "lr":0.0100 ,  "loss":0.8960 ,  "loss_classifier":0.0784 ,  "loss_box_reg":0.0326 ,  "loss_mask":0.6608 ,  "loss_objectness":0.0534 ,  "loss_rpn_box_reg":0.0162 , "time":0.7539, "data":0.1020, "GPUMemory":3156 }
{
    
     "epoch":5, "imgindex":60, "imgcount":63,  "lr":0.0100 ,  "loss":0.8618 ,  "loss_classifier":0.0936 ,  "loss_box_reg":0.0499 ,  "loss_mask":0.6044 ,  "loss_objectness":0.0594 ,  "loss_rpn_box_reg":0.0152 , "time":0.7365, "data":0.0846, "GPUMemory":3156 }
{
    
     "epoch":5, "imgindex":62, "imgcount":63,  "lr":0.0100 ,  "loss":0.9147 ,  "loss_classifier":0.0936 ,  "loss_box_reg":0.0499 ,  "loss_mask":0.6383 ,  "loss_objectness":0.0447 ,  "loss_rpn_box_reg":0.0152 , "time":0.6980, "data":0.0501, "GPUMemory":3156 }
{
    
     "epoch":6, "imgindex":0, "imgcount":63,  "lr":0.0100 ,  "loss":1.5819 ,  "loss_classifier":0.2886 ,  "loss_box_reg":0.2014 ,  "loss_mask":0.9627 ,  "loss_objectness":0.1051 ,  "loss_rpn_box_reg":0.0240 , "time":13.2378, "data":12.5113, "GPUMemory":3156 }
{
    
     "epoch":6, "imgindex":20, "imgcount":63,  "lr":0.0100 ,  "loss":0.8689 ,  "loss_classifier":0.0960 ,  "loss_box_reg":0.0419 ,  "loss_mask":0.6413 ,  "loss_objectness":0.0611 ,  "loss_rpn_box_reg":0.0136 , "time":0.7089, "data":0.0535, "GPUMemory":3156 }
{
    
     "epoch":6, "imgindex":40, "imgcount":63,  "lr":0.0100 ,  "loss":0.7758 ,  "loss_classifier":0.0517 ,  "loss_box_reg":0.0251 ,  "loss_mask":0.5768 ,  "loss_objectness":0.0629 ,  "loss_rpn_box_reg":0.0119 , "time":0.9487, "data":0.2802, "GPUMemory":3156 }
{
    
     "epoch":6, "imgindex":60, "imgcount":63,  "lr":0.0100 ,  "loss":0.8633 ,  "loss_classifier":0.0808 ,  "loss_box_reg":0.0355 ,  "loss_mask":0.6135 ,  "loss_objectness":0.0591 ,  "loss_rpn_box_reg":0.0149 , "time":0.6702, "data":0.0343, "GPUMemory":3156 }
{
    
     "epoch":6, "imgindex":62, "imgcount":63,  "lr":0.0100 ,  "loss":0.8633 ,  "loss_classifier":0.0808 ,  "loss_box_reg":0.0355 ,  "loss_mask":0.6135 ,  "loss_objectness":0.0591 ,  "loss_rpn_box_reg":0.0149 , "time":0.6635, "data":0.0342, "GPUMemory":3156 }
{
    
     "epoch":7, "imgindex":0, "imgcount":63,  "lr":0.0100 ,  "loss":0.2361 ,  "loss_classifier":0.0167 ,  "loss_box_reg":0.0002 ,  "loss_mask":0.1616 ,  "loss_objectness":0.0511 ,  "loss_rpn_box_reg":0.0065 , "time":12.7431, "data":12.1155, "GPUMemory":3156 }
{
    
     "epoch":7, "imgindex":20, "imgcount":63,  "lr":0.0100 ,  "loss":0.8250 ,  "loss_classifier":0.0889 ,  "loss_box_reg":0.0388 ,  "loss_mask":0.6147 ,  "loss_objectness":0.0535 ,  "loss_rpn_box_reg":0.0140 , "time":0.8481, "data":0.2082, "GPUMemory":3156 }
{
    
     "epoch":7, "imgindex":40, "imgcount":63,  "lr":0.0100 ,  "loss":0.8195 ,  "loss_classifier":0.0845 ,  "loss_box_reg":0.0551 ,  "loss_mask":0.5938 ,  "loss_objectness":0.0565 ,  "loss_rpn_box_reg":0.0155 , "time":0.8044, "data":0.1416, "GPUMemory":3156 }
{
    
     "epoch":7, "imgindex":60, "imgcount":63,  "lr":0.0100 ,  "loss":0.7695 ,  "loss_classifier":0.0654 ,  "loss_box_reg":0.0317 ,  "loss_mask":0.5683 ,  "loss_objectness":0.0475 ,  "loss_rpn_box_reg":0.0105 , "time":0.7677, "data":0.1099, "GPUMemory":3156 }
{
    
     "epoch":7, "imgindex":62, "imgcount":63,  "lr":0.0100 ,  "loss":0.7389 ,  "loss_classifier":0.0654 ,  "loss_box_reg":0.0317 ,  "loss_mask":0.5634 ,  "loss_objectness":0.0472 ,  "loss_rpn_box_reg":0.0127 , "time":0.7940, "data":0.1415, "GPUMemory":3156 }
{
    
     "epoch":8, "imgindex":0, "imgcount":63,  "lr":0.0100 ,  "loss":0.5342 ,  "loss_classifier":0.0315 ,  "loss_box_reg":0.0125 ,  "loss_mask":0.4456 ,  "loss_objectness":0.0313 ,  "loss_rpn_box_reg":0.0133 , "time":13.4497, "data":12.7742, "GPUMemory":3156 }

刚开始是这样的.代码


# 读取文本文件, 
# 1. 自动去除utf8 BOM 头
# 2. 统一 \r\n 转成 \n
def ReadAllTextFromFile(str,encoding='utf-8'):
    try: 
        with open(str,'r+',encoding=encoding) as f:
            s = f.read().replace('\r\n','\n')
            if ord(s[0]) == 65279:# 65279 对应的是 \xfeff 是BOM头.所以去掉
                return s[1:]
            if ord(s[0]) == 65534:# 65534 对应的是 \xfffe 是BOM头.所以去掉
                return s[1:]
            if ord(s[0]) == 4294836224:# 4294836224 对应的是 \xfffe0000 是BOM头.所以去掉
                return s[1:]
            return s
    except Exception as ex:
        print(ex)
        return ""

# 取得所有的行数, splitlines 不带换行符 
def ReadAllLinesFromFile(str,encoding='utf-8'):
    txt = ReadAllTextFromFile(str,encoding)
    return txt.splitlines()

# 根据results.txt 绘制关键参数results.png图, 
def plot_resultsimage(save_dir=''):  # from utils.general import *; plot_results()
    # Plot training 'results*.txt' as seen in https://github.com/ultralytics/yolov5#reproduce-our-training
    
    fig, ax = plt.subplots(2, 4, figsize=(12, 6))
    ax = ax.ravel()
    labels = ['lr','loss','loss_classifier','loss_box_reg','loss_mask','loss_objectness','loss_rpn_box_reg']
    files = glob.glob(str(Path(save_dir) / 'results*.txt'))
    for fi, f in enumerate(files):
        # try:
            # results = np.loadtxt(f, usecols=[2, 3, 4, 8, 9, 12, 13, 14, 10, 11], ndmin=2).T
            lines = ReadAllLinesFromFile(f)
            results = []
            for txtrow in lines:
              results.append(json.loads(txtrow))

            for colindex, label in enumerate(labels): 
              for rowindex, row in enumerate(results):
                  ax[colindex].plot(rowindex, row[label], linestyle='--', marker='.', label=label, linewidth=2, markersize=2)
                  ax[colindex].set_title(label)
            
              
        # except Exception as e:
        #     print(str(e))

    # fig.tight_layout()
    # ax[1].legend()
    fig.savefig(Path(save_dir) / 'results.png', dpi=200)
    # fig.show()
    pass

得到结果是这样的.没有折线.百思不得其解
在这里插入图片描述
而且看上去好像是不对的.后来仔细的看了很多次, 改成下面的代码就可以了

正确的绘制折线图的方法


import matplotlib
import matplotlib.pyplot as plt
import glob
from pathlib import Path
import numpy as np
import json

# 读取文本文件, 
# 1. 自动去除utf8 BOM 头
# 2. 统一 \r\n 转成 \n
def ReadAllTextFromFile(str,encoding='utf-8'):
    try: 
        with open(str,'r+',encoding=encoding) as f:
            s = f.read().replace('\r\n','\n')
            if ord(s[0]) == 65279:# 65279 对应的是 \xfeff 是BOM头.所以去掉
                return s[1:]
            if ord(s[0]) == 65534:# 65534 对应的是 \xfffe 是BOM头.所以去掉
                return s[1:]
            if ord(s[0]) == 4294836224:# 4294836224 对应的是 \xfffe0000 是BOM头.所以去掉
                return s[1:]
            return s
    except Exception as ex:
        print(ex)
        return ""

# 取得所有的行数, splitlines 不带换行符 
def ReadAllLinesFromFile(str,encoding='utf-8'):
    txt = ReadAllTextFromFile(str,encoding)
    return txt.splitlines()
    
def plot_resultsimage(save_dir=''):  # from utils.general import *; plot_results()
    # Plot training 'results*.txt' as seen in https://github.com/ultralytics/yolov5#reproduce-our-training
    
    fig, ax = plt.subplots(2, 4, figsize=(12, 6))
    ax = ax.ravel()
    labels = ['lr','loss','loss_classifier','loss_box_reg','loss_mask','loss_objectness','loss_rpn_box_reg']
    files = glob.glob(str(Path(save_dir) / 'results*.txt'))
    for fi, f in enumerate(files):
        # try:
            # results = np.loadtxt(f, usecols=[2, 3, 4, 8, 9, 12, 13, 14, 10, 11], ndmin=2).T
            lines = ReadAllLinesFromFile(f)
            results = []
            for txtrow in lines:
              results.append(json.loads(txtrow)) 


            for colindex, label in enumerate(labels): 
                # for rowindex, row in enumerate(results):
                coldatas = [row[label] for row in results] 
                ax[colindex].plot(coldatas,  marker='.', label=label, linewidth=1, markersize=3)
                ax[colindex].set_title(label)
            
              
        # except Exception as e:
        #     print(str(e))

    # fig.tight_layout()
    # ax[1].legend()
    fig.savefig(Path(save_dir) / 'results.png', dpi=200)
    # fig.show()
    pass

在这里插入图片描述
这样的结果就正常了.
第一种代码在输出的时候, 依次循环每个坐标图, 每次画一个点,导致它无法知道上一个点的位置, 所以也就无法连续性的绘制折线,
只要一次性给每个子图投入所有的数据点即可绘制出折线, 而且第二次再调用的时候, 线的颜色会自动变化.
主要的代码变化在 coldatas = [row[label] for row in results] 这一行.

会猜真的是太重要了…

猜你喜欢

转载自blog.csdn.net/phker/article/details/109623851
今日推荐
openKylin 社区生态委员会第六次会议圆满召开
阿里云正式发布通义千问 2.5
Python 3.13 发布首个 Beta：实验性自由线程模式和 JIT、改进交互式解释器
Stack Overflow 拿我的代码去训练 AI 大模型，还封了我的账号
Pop!_OS 的 COSMIC 桌面完成 App Store 上架工作
报告：Django 仍然是 74% 开发者的首选
《2024 年一季度互联网投融资运行情况》研究报告
15 年前上了“FFmpeg 耻辱柱”，今天他还得谢谢咱——腾讯QQPlayer一雪前耻？
TIOBE 5 月榜单：Fortran “复活”进入 Top 10
GCC 14.1 发布
面壁智能发布 Eurux-8x22B 开源大模型 —— 堪称「理科状元」
开源日报 | 谷歌扶持鸿蒙上位；开源Rabbit R1；Docker加持的安卓手机；微软的焦虑和野心；海尔电器把开放平台关了
周排行
计算机组成与设计（七）—— 除法器
Integer Approximation(分治+枚举)
大话数据库索引
windows10系统JDK的配置及下载地址
mysql实现秒值转换中原六仔平台搭建
Codeforces Round #556 (Div. 1)
百练1064 网线主管
Codeforces 995F Cowmpany Cowmpensation
子集生成之增量构造法，位向量法，二进制法
ERROR: cmd.exe failed with args /c "/APK\gradle\rungradle.bat...
每日归档
更多
2024-05-10(38)
2024-05-09(35)
2024-05-08(42)
2024-05-07(14)
2024-05-06(40)
2024-05-05(0)
2024-05-04(7)
2024-05-03(19)
2024-05-02(0)
2024-05-01(4)

代码天地 © 2018 codetd.com

境外服务器   最新电视剧2022