2地点所要時間分析¶

2地点間の流動アドレスについて、時間間隔(departureとarrival)の分布を調べてみる。

id	lat	lon	地点名
kofu2	35.658933	138.571024	三枝豆店
kofu3	35.660988	138.571165	風月堂
kofu4	35.670632	138.565893	永田楽器
kofu5	35.662019	138.571818	文化のるつぼ Hechima
kofu7	35.666474	138.567104	ライフインナカゴミ
kofu8	35.669087	138.566179	オスカー本社・朝日店
kofu9	35.663483	138.568223	防災新館1F
kofu10	35.660508	138.571015	河野スポーツ
kofu11	35.65898	138.569654	内藤セイビドー眼鏡店
kofu12	35.665457	138.567812	立ち食い焼肉鷹の
kofu13	35.659908	138.571535	ブラザー（文房具）
kofu14	35.666835	138.571457	甲州夢小路
kofu16	35.662705	138.570433	甲府クラフトラボ
kofu17	35.661394	138.56881	ダン珈琲店
kofu19	35.660883	138.570182	きぬや
kofu20	35.66005	138.56915	そば処奥義
kofu21	35.671693	138.566729	玉屋
kofu22	35.665733	138.568327	奥藤本店
kofu23	35.665305	138.567282	六曜館珈琲店本店
kofu24	35.660092	138.570098	ルパンザバールワイン酒場前

import pandas as pd
import numpy as np
import matplotlib
from datetime import datetime, timedelta

sdate = "20181201" # 日にち範囲指定
edate = "20181207"
base_filename = "/home/raspimngr/wifi_csv/summary_kofu/traveler/travelerCount_kofu_"
day = datetime.strptime(sdate, "%Y%m%d")
last_day = datetime.strptime(edate, "%Y%m%d")

src_data = pd.DataFrame(columns=["addr","start","dep_date","goal",
                                 "arrival_date", "local_bit"])
while day <= last_day:
    day_str = datetime.strftime(day, "%Y%m%d")
    tr_data = pd.read_csv(base_filename + day_str + ".csv",
                          sep=",",
                          names=("addr","start","dep_date",
                                 "goal", "arrival_date", "local_bit"))
    src_data =pd.concat([src_data,tr_data], ignore_index=True)
    # src_data = src_data.append(tr_data, ignore_index=True) # 上記と同じ
    day += timedelta(days=1)
    
src_data.info()
# src_data.head()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 1014016 entries, 0 to 1014015
Data columns (total 6 columns):
addr            1014016 non-null object
start           1014016 non-null object
dep_date        1014016 non-null object
goal            1014016 non-null object
arrival_date    1014016 non-null object
local_bit       1014016 non-null object
dtypes: object(6)
memory usage: 46.4+ MB

from datetime import datetime as dt

src_data["time_delta"] = (pd.to_datetime(src_data['arrival_date']) - 
                        pd.to_datetime(src_data['dep_date']))
src_data["total_seconds"] = [i.total_seconds() for i in src_data.time_delta]
# 所要時間を分刻みとして追加
src_data["total_min"] = (src_data["total_seconds"]/60).astype(int)

src_data.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 1014016 entries, 0 to 1014015
Data columns (total 9 columns):
addr             1014016 non-null object
start            1014016 non-null object
dep_date         1014016 non-null object
goal             1014016 non-null object
arrival_date     1014016 non-null object
local_bit        1014016 non-null object
time_delta       1014016 non-null timedelta64[ns]
total_seconds    1014016 non-null float64
total_min        1014016 non-null int64
dtypes: float64(1), int64(1), object(6), timedelta64[ns](1)
memory usage: 69.6+ MB

# 9: 防災会館 17:ダン珈琲店
hist_dt = src_data[(src_data["start"]==9) &
                   (src_data["goal"]==17) &
                   (src_data["local_bit"] == 0)]
hist_dt.tail()

bin_list = [0,1,2,3,5,10,30,60,600]
pd.cut(hist_dt["total_min"],bin_list, right=False).value_counts(sort=True)

[1, 2)       587
[60, 600)    543
[0, 1)       505
[2, 3)       442
[3, 5)       409
[10, 30)     252
[5, 10)      222
[30, 60)     129
Name: total_min, dtype: int64

bin_list = [0,1,2,3,5,10,30,60,600]
bin_label = ["0-1", "1-2","2-3","3-5", "5-10", "10-30","30-60", "60<"]
data = pd.cut(hist_dt["total_min"],bins=bin_list, labels=bin_label,
              right=False).value_counts(sort=False)
ax = data.plot.bar()
ax.set_title("防災会館 - ダン珈琲店 観測時刻間隔(12/01-12/07)", size=14)
ax.set_xlabel("minutes", size=14)
ax.set_ylabel("number of global addresses", size=14)

Text(0,0.5,'number of global addresses')

防災会館とダン珈琲店の距離はGoogle Mapによれば約300m(徒歩約4分)

１分以内は車
1分～3分は車or自転車

とすると、車と自転車・バイクはほぼ同程度ではないか。

	addr	start	dep_date	goal	arrival_date	time_delta	total_seconds	total_min
1011964	fc:08:4a:5c:a2:01	9	2018-12-07 07:27:10	17	2018-12-07 07:28:26	00:01:16	76.0	1
1011970	fc:08:4a:5c:a3:f3	9	2018-12-07 16:45:06	17	2018-12-07 16:47:02	00:01:56	116.0	1
1011993	fc:08:4a:5c:bc:3c	9	2018-12-07 18:05:23	17	2018-12-07 18:07:25	00:02:02	122.0	2
1012159	fc:62:b9:d3:50:d8	9	2018-12-07 15:57:13	17	2018-12-07 15:58:47	00:01:34	94.0	1
1012336	fc:d8:48:29:4f:53	9	2018-12-07 15:22:36	17	2018-12-07 15:34:41	00:12:05	725.0	12