خوشه بندی دیتاهای حجیم به جهت نمایش در نقشه

				
					def date_diff_in_Seconds(dt1, dt2):
    # Calculate the time difference between dt2 and dt1
    # Step 2: Extract milliseconds separately and add them to the parsed datetime object
    timedelta = datetime.strptime(dt2[:-4], '%m/%d/%Y %H:%M:%S') - datetime.strptime(dt1[:-4], '%m/%d/%Y %H:%M:%S')
    # Return the total time difference in seconds
    return timedelta.days * 24 * 3600 + timedelta.seconds

ef get_2GLocked_IDLE_cluster_fromExcel():
    KPI_gsm = pd.read_excel("Original.xlsx", sheet_name='GSM_KPI')
    lstGSMRanges = []
    base = -150
    offset = 5
    cluster_samples = 10
    cluster_Time = 20
    min = base
    max = base + 5
    while min < 0:
        crange = KPI_gsm.query("`RxLevel` >= @min and `RxLevel` < @max")
        if len(crange.index) >0:
            lstGSMRanges.append(crange)
        min = max
        max += 5
    clusters = pd.DataFrame(columns=['DateTime', 'Latitude', 'Longitude','CellID','RxLevel'])
    c_cluster = pd.DataFrame(columns=['DateTime', 'Latitude', 'Longitude','CellID','RxLevel'])
    cluster_time=10

    for gsmrange in lstGSMRanges:
        for index, row in gsmrange.iterrows():
            c_cluster=c_cluster.append({'DateTime':row['Date Time'],'Latitude':row['Latitude'],'Longitude':row['Longitude'],'CellID':row['Cell-ID'],'RxLevel':row['RxLevel']}, ignore_index=True)
            if len(c_cluster.index) > 1:
                if postfunc.date_diff_in_Seconds(c_cluster.iloc[0]['DateTime'],c_cluster.iloc[len(c_cluster.index)-1]['DateTime']) >= cluster_time or len(c_cluster.index)==5:
                    clusters = clusters.append({'DateTime':c_cluster.loc[0]['DateTime'],'Latitude':c_cluster.loc[0]['Latitude'],'Longitude':c_cluster.loc[0]['Longitude'],'CellID':c_cluster.loc[0]['CellID'],'RxLevel':c_cluster['RxLevel'].mean()}, ignore_index=True)
                    c_cluster = c_cluster.iloc[0:0]
    clusters.to_csv('clusters.csv', index=False)