ADD file via upload

analysis.py

@@ -0,0 +1,121 @@
+import pandas as pd
+import matplotlib.pyplot as plt
+import numpy as np
+from wordcloud import WordCloud
+import jieba
+import cv2 as cv
+from pylab import mpl
+import os
+from matplotlib.font_manager import _rebuild
+_rebuild() #reload一下
+
+# 设置显示中文字体
+mpl.rcParams["font.sans-serif"] = ["FangSong"]
+# 设置正常显示符号
+mpl.rcParams["axes.unicode_minus"] = False
+
+def prepressingData(data1):
+    data = data1.copy()
+    zancheng = data['赞成数']
+    fandui = data['反对数']
+    huifu = data['回复数']
+    pingjia = data['评价']
+    star = []
+    for i in range(0,len(zancheng)):
+        zancheng[i] = str(zancheng[i]).replace('\n','')
+        fandui[i] = str(fandui[i]).replace('\n','')
+        huifu[i] = str(huifu[i]).replace('回应','')
+        if pingjia[i]=='力荐':
+            star.append(5)
+        elif pingjia[i]=='推荐':
+            star.append(4)
+        elif pingjia[i]=='还行':
+            star.append(3)
+        elif pingjia[i]=='较差':
+            star.append(2)
+        elif pingjia[i]=='很差':
+            star.append(1)
+        else:
+            star.append(0)
+    data['评价得分'] = pd.Series(star)
+    data = data.drop(axis=1,columns=['评论主体'])
+#     删除列，后要赋值过去才算成功
+    return data
+
+def cvtInt(data1):
+    data = data1.copy()
+    data['赞成数'] = pd.to_numeric(data1['赞成数'],errors='ignore')
+    data['赞成数'] = data['赞成数'].fillna(0)
+    data['反对数'] = pd.to_numeric(data1['反对数'],errors='ignore')
+    data['反对数'] = data['反对数'].fillna(0)
+    data['回复数'] = pd.to_numeric(data1['回复数'],errors='ignore')
+    return data
+
+def dropUselessTime(data):
+    data1 = data.copy()
+    index = data1[data1['评论时间']=='no star'].index
+    data1 = data1.drop(axis=0,index=index)
+    return data1
+
+
+def plotTimeWithData(data,path):
+    data1 = data.copy()
+    cur = data1.groupby('hour').count()
+    plt.figure(figsize=(8, 6))
+    plt.plot(cur.index, cur['month'])
+    plt.xlabel(u'时间/hour', size=23)
+    plt.ylabel(u'评论数量', size=23)
+    plt.title(u'评论数量随时间段的变化', size=23)
+    xticks = list(range(0, 24))  # 这里设置的是x轴点的位置
+    plt.xticks(xticks)
+    plt.savefig(path + '评论数量随时间段的变化.png')
+
+    cur = data1.groupby('hour').mean()
+    plt.figure(figsize=(8, 6))
+    plt.plot(cur.index, cur['评价得分'])
+    plt.xlabel('时间/hour', size=23)
+    plt.ylabel('评论分均值', size=23)
+    plt.title('评论均值随时间段的变化', size=23)
+    xticks = list(range(0, 24))  # 这里设置的是x轴点的位置
+    plt.xticks(xticks)
+    plt.savefig(path + '评论均值随时间段的变化.png')
+
+
+def save_pics(name,path):
+    pic_path = path + name + '.jpg'
+    print(pic_path)
+    background_image = cv.imread(pic_path)
+    if not os.path.exists(path):
+        os.makedirs(path)
+    filename = name + '.csv'
+    yingping = pd.read_csv(path + '\\' + filename)
+    all_content = ''
+    for i in range(0,len(yingping['评论标题'])):
+        all_content= all_content+yingping['评论标题'][i]
+    cut_text = " ".join(jieba.cut(all_content))
+    wordcloud = WordCloud(font_path="C:/Windows/Fonts/SimHei.ttf",mask=background_image,background_color='white').generate(cut_text)
+    plt.figure(figsize=(10,10))
+    plt.imshow(wordcloud,interpolation="bilinear")
+    plt.axis("off")
+    # plt.show()
+    plt.savefig(path + '词云.png')
+
+    yingping = prepressingData(yingping)
+    yingping.head(50)
+    yingping.isnull().any()
+
+    # 画图开始
+    plt.rcParams['font.sans-serif'] = ['SimHei']
+    plt.figure(figsize=(10, 8))
+    plt.hist(yingping['评价得分'], bins=15)
+    plt.xlabel('星级（0表示没有评级）', size=14)
+    plt.ylabel('人数', size=14)
+    # plt.show()
+    plt.savefig(path + '各星级的人数统计.png')
+
+    yingping = cvtInt(yingping)
+    yingping = dropUselessTime(yingping)
+    yingping.index = yingping['评论时间']
+    yingping['month'] = pd.to_datetime(yingping['评论时间']).dt.month
+    yingping['hour'] = pd.to_datetime(yingping['评论时间']).dt.hour
+    plotTimeWithData(yingping,path)