예제 #1
0
 def calculateLambda(self, data1, data2, num_pois):
     distance = []
     data_viz = dataViz.DataViz()
     for i, poi in data2.iterrows():
         distance.append(self.algo.calculateDis(poi, data1))
     coordonnee = self.drawPoint(distance, num_pois)
     coordonnee.append(data1['product_name'])
     return coordonnee
예제 #2
0
        'water-hardness_100g', 'choline_100g', 'phylloquinone_100g',
        'beta-glucan_100g', 'inositol_100g', 'carnitine_100g'
    ]
    df = dataCenter.cleanData(df, colslist)
    cols = df.columns.values.tolist()
    df.info()

    clean_brand = cleanBrand.CleanBrand()
    dfBrand1 = clean_brand.getBrandsClean()
    df['brand1'] = dfBrand1

    df['brand1'] = df['brand1'].fillna(value='NoBrand', axis=0)

    # 计算两条记录的2gram的余弦值
    algo = algoDis.AlgoDis()
    data_Viz = dataViz.DataViz()

    # 并行!!!!
    num_pois = 5
    num_divide_df = 100

    # 开始选取参考点
    timeStartKmeans = time.localtime()
    print('choose POIs: ' + str(timeStartKmeans))
    list_pois, pois = dataCenter.selectPOIs(df.iloc[1:5000], num_pois)
    # list_pois, pois = dataCenter.selectPOIsRandom(num_pois,df)
    list_pois_labels = []
    for index in list_pois:
        list_pois_labels.append(index)
    df_copy = df
    df_copy = df_copy.drop(list_pois, axis=0)