notebooks/atlas_dataprep.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## RIPE Atlas Data Preparation for Latency Privacy Study\n",
    "\n",
    "This notebook contains code for downloading anchoring traceroute and ping measurements from RIPE Atlas, and dumping them into an HDF5 datastore for later analysis.\n",
    "\n",
    "The output consists of three dataframes:\n",
    "\n",
    "- `rtt_df` contains the following columns:\n",
    "  - `time`: timestamp of measurement sample\n",
    "  - `pid`: probe ID\n",
    "  - `aid`: anchor ID\n",
    "  - `af`: address family (4 or 6)\n",
    "  - `proto`: protocol used (\"ICMP_PING\" or \"ICMP_TR\" for now)\n",
    "  - `hopcount`: total count of hops in measurement (0 = ping)\n",
    "  - `hop`: hop of this measurement (for n-k traceroute)\n",
    "  - `rtt`: round trip time in microseconds\n",
    "\n",
    "- `probe_df` contains the following columns:\n",
    "  - `pid`: probe ID\n",
    "  - `version`: probe software version (-1 = anchor, 0 = unknown)\n",
    "  - `nat`: true if probe NATted\n",
    "  - `ip4`: IPv4 address\n",
    "  - `ip6`: IPv6 address\n",
    "  - `asn4`: IPv4 BGP ASN\n",
    "  - `asn6`: IPv6 BGP ASN\n",
    "  - `cc`: country code (two-letter 3166)\n",
    "  - `lon`: longitude \n",
    "  - `lat`: latitude\n",
    "\n",
    "- `anchor_df` contains the following columns:\n",
    "  - `aid`: anchor ID\n",
    "  - `name`: anchor name, i.e. FQDN without .anchors.atlas.ripe.net \n",
    "  - `pid`: probe ID of colocated probe\n",
    "  - `ip4`: IPv4 address\n",
    "  - `ip6`: IPv6 address\n",
    "  - `asn4`: IPv4 BGP ASN\n",
    "  - `asn6`: IPv6 BGP ASN\n",
    "  - `cc`: country code (two-letter 3166)\n",
    "  - `lon`: longitude \n",
    "  - `lat`: latitude\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import dateutil.parser as dp\n",
    "import requests\n",
    "import json\n",
    "import os.path\n",
    "\n",
    "from collections import namedtuple\n",
    "\n",
    "import warnings\n",
    "warnings.filterwarnings(\"ignore\") "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Configuration\n",
    "\n",
    "Fill in configuration values here, the run all cells in the notebook to produce an HDF5 file to use in subsequent analysis notebooks."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Local path to cache Atlas measurement data in\n",
    "ATLAS_CACHE_PATH=\"atlas_cache_20171002\"\n",
    "\n",
    "# Filename for HDF5 datastore\n",
    "STORE_PATH=\"atlas_20171002.hdf5\"\n",
    "\n",
    "# Directory for plots\n",
    "PLOT_PATH=\"../plots\"\n",
    "\n",
    "# Start and end time for Atlas measurement retrieval\n",
    "START_TIME = int(dp.parse(\"2017-10-02T00:00:00Z\").timestamp())\n",
    "STOP_TIME = int(dp.parse(\"2017-10-03T00:00:00Z\").timestamp())\n",
    "\n",
    "# Names of anchors we'd like to study\n",
    "ANCHOR_NAMES_WE_LIKE = [\n",
    "    \"ar-bue-as4270\",   # Buenos Aires, Argentina\n",
    "    \"at-vie-as1120\",   # Vienna, Austria\n",
    "    \"au-mel-as38796\",  # Melbourne, Austraila\n",
    "    \"au-bne-as4608-2\", # Brisbane, Australia\n",
    "    \"bd-dac-as24122\",  # Dacca, Bangladesh\n",
    "    \"bf-oua-as37577\",  # Ouagadougou, Burkina Faso\n",
    "    \"bg-sof-as8866\",   # Sofia, Bulgaria\n",
    "    \"br-sao-as22548\",  # Sao Paulo, Brasil\n",
    "    \"ca-mtr-as852\",    # Montreal, Canada\n",
    "    \"ca-wnp-as18451\",  # Winnepeg, Canada\n",
    "    \"ch-gtg-as20612\",  # Glattbrug, Switzerland\n",
    "    \"ch-zrh-as559\",    # Zurich, Switzerland\n",
    "    \"cz-brq-as197451\", # Brno, Czechia\n",
    "    \"de-fra-as8763\",   # Frankfurt, Germany\n",
    "    \"de-ham-as201709\", # Hamburg, Germany\n",
    "    \"de-muc-as5539\",   # Munich, Germany\n",
    "    \"ee-tll-as51349\",  # Talinn, Estonia\n",
    "    \"es-bcn-as13041\",  # Barcelona, Spain\n",
    "    \"fr-par-as1307\",   # Paris, France\n",
    "    \"gr-ath-as5408\",   # Athens, Greece\n",
    "    \"hk-hkg-as43996\",  # Hong Kong SAR, China\n",
    "    \"hu-bud-as12303\",  # Budapest, Hungary\n",
    "    \"id-jkt-as10208\",  # Jakarta, Indonesia\n",
    "    \"ie-dub-as1213\",   # Dublin, Ireland\n",
    "    \"in-bom-as33480\",  # Mumbai, India\n",
    "    \"it-trn-as12779\",  # Turin, Italy\n",
    "    \"jp-tyo-as2500\",   # Tokyo, Japan\n",
    "    \"ke-nbo-as37578\",  # Nairobi, Kenya\n",
    "    \"kz-ala-as21299\",  # Almaty, Kazakhstan\n",
    "    \"nl-ams-as3333\",   # Amsterdam, Netherlands\n",
    "    \"nl-haa-as201682\", # Haarlen, Netherlands\n",
    "    \"nz-wlg-as9834\",   # Wellington, New Zealand\n",
    "    \"qa-doh-as8781\",   # Doha, Qatar\n",
    "    \"ru-mow-as15835\",  # Moscow, Russia\n",
    "    \"se-sto-as8674\",   # Stockholm, Sweden\n",
    "    \"uk-lon-as5607\",   # London, England\n",
    "    \"us-dal-as2914\",   # Dallas, USA\n",
    "    \"us-den-as7922\",   # Denver, USA\n",
    "    \"us-mia-as33280\",  # Miami, USA\n",
    "    \"us-sjc-as22300\",  # San Jose, USA\n",
    "]\n",
    "\n",
    "# Protocols to extract latency information from\n",
    "PROTOCOLS_WE_LIKE = [\n",
    "    \"ICMP_TR\",      # Traceroute\n",
    "    #\"ICMP_PING\",   # Ping\n",
    "    #\"TCP_HTTP\"     # HTTP\n",
    "]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "40"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(ANCHOR_NAMES_WE_LIKE)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Code\n",
    "\n",
    "No user-serviceable parts inside. :)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Probe metadata"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Extract probe information from files\n",
    "AtlasProbe = namedtuple(\"AtlasProbe\",\n",
    "           (\"pid\", \"version\", \"nat\", \"ip4\", \"ip6\", \"asn4\", \"asn6\", \"cc\", \"lon\", \"lat\"))\n",
    "\n",
    "def extract_atlas_probe(pobj):\n",
    "    \n",
    "    if \"address_v4\" in pobj and pobj['address_v4'] is not None:\n",
    "        ip4 = pobj[\"address_v4\"]\n",
    "    elif \"prefix_v4\" in pobj:\n",
    "        ip4 = pobj[\"prefix_v4\"]\n",
    "    else:\n",
    "        ip4 = None\n",
    "\n",
    "    if \"address_v6\" in pobj and pobj['address_v6'] is not None:\n",
    "        ip6 = pobj[\"address_v6\"]\n",
    "    elif \"prefix_v4\" in pobj:\n",
    "        ip6 = pobj[\"prefix_v6\"]\n",
    "    else:\n",
    "        ip6 = None\n",
    "\n",
    "    if \"asn_v4\" in pobj:\n",
    "        asn4 = pobj[\"asn_v4\"]\n",
    "    else:\n",
    "        asn4 = None\n",
    "\n",
    "    if \"asn_v6\" in pobj:\n",
    "        asn6 = pobj[\"asn_v6\"]\n",
    "    else:\n",
    "        asn6 = None\n",
    "\n",
    "    if \"tags\" in pobj:\n",
    "        if len(pobj['tags']) > 0 and isinstance(pobj['tags'][0], dict):\n",
    "            alltags = [tag['slug'] for tag in pobj['tags']]\n",
    "        else:\n",
    "            alltags = pobj['tags']\n",
    "\n",
    "        if \"system-v1\" in alltags:\n",
    "            version = 1\n",
    "        elif \"system-v2\" in alltags:\n",
    "            version = 2\n",
    "        elif \"system-v3\" in alltags:\n",
    "            version = 3\n",
    "        elif \"system-anchor\" in alltags:\n",
    "            version = -1\n",
    "        else:\n",
    "            version = 0 # unknown version - never connected?\n",
    "\n",
    "        nat = \"nat\" in alltags\n",
    "    else:\n",
    "        version = None\n",
    "        nat = None\n",
    "        \n",
    "    # Short circuit: never connected means don't load\n",
    "    if \"status\" in pobj and pobj['status'] == 0:\n",
    "        version = 0\n",
    "    \n",
    "    if \"geometry\" in pobj and \"coordinates\" in pobj['geometry']:\n",
    "        (lon, lat) = pobj['geometry']['coordinates']\n",
    "    elif \"longitude\" in pobj and \"latitude\" in pobj:\n",
    "        lon = pobj['longitude']\n",
    "        lat = pobj['latitude']\n",
    "    else:\n",
    "        lon = None\n",
    "        lat = None\n",
    "\n",
    "    return AtlasProbe(pobj[\"id\"], version, nat, ip4, ip6, asn4, asn6,\n",
    "                      pobj[\"country_code\"], lon, lat)\n",
    "\n",
    "def probe_dataframe_from_file(filename):\n",
    "    data = []\n",
    "    \n",
    "    # make a giant array\n",
    "    with open(filename) as stream:\n",
    "        all_probes = json.loads(stream.read())\n",
    "        for pobj in all_probes[\"objects\"]:\n",
    "            data.append(extract_atlas_probe(pobj))\n",
    "\n",
    "    # create a dataframe from it\n",
    "    df = pd.DataFrame(data, columns=AtlasProbe._fields)\n",
    "    \n",
    "    # indexed by probe ID\n",
    "    df.index = df['pid']\n",
    "    del(df['pid'])\n",
    "    \n",
    "    # and return it\n",
    "    return df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "  % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current\n",
      "                                 Dload  Upload   Total   Spent    Left  Speed\n",
      "100  887k  100  887k    0     0  1177k      0 --:--:-- --:--:-- --:--:-- 1176k\n"
     ]
    }
   ],
   "source": [
    "! curl http://ftp.ripe.net/ripe/atlas/probes/archive/meta-latest | bunzip2 > all_probes.json"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "probe_df = probe_dataframe_from_file(\"all_probes.json\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Anchor metadata"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Extract anchor information from API\n",
    "AtlasAnchor = namedtuple(\"AtlasAnchor\",\n",
    "           (\"aid\", \"name\", \"pid\", \"ip4\", \"ip6\", \"asn4\", \"asn6\", \"cc\", \"lon\", \"lat\"))\n",
    "\n",
    "def extract_atlas_anchor(aobj):\n",
    "    \n",
    "    if \"id\" in aobj:\n",
    "        aid = int(aobj[\"id\"])\n",
    "    else:\n",
    "        aid = None\n",
    "\n",
    "    if \"fqdn\" in aobj:\n",
    "        name = aobj[\"fqdn\"]\n",
    "    else:\n",
    "        name = None\n",
    "    \n",
    "    if name is not None and name.endswith(\".anchors.atlas.ripe.net\"):\n",
    "        name = name[:-23]\n",
    "        \n",
    "    if \"probe\" in aobj:\n",
    "        pid = int(aobj[\"probe\"])\n",
    "    else:\n",
    "        pid = None\n",
    "    \n",
    "    if \"ip_v4\" in aobj:\n",
    "        ip4 = aobj[\"ip_v4\"]\n",
    "    else:\n",
    "        ip4 = None\n",
    "    \n",
    "    if \"ip_v6\" in aobj:\n",
    "        ip6 = aobj[\"ip_v6\"]\n",
    "    else:\n",
    "        ip6 = None\n",
    "\n",
    "    if \"as_v4\" in aobj and aobj['as_v4'] is not None:\n",
    "        asn4 = int(aobj[\"as_v4\"])\n",
    "    else:\n",
    "        asn4 = None\n",
    "\n",
    "    if \"as_v6\" in aobj and aobj['as_v6'] is not None:\n",
    "        asn6 = int(aobj[\"as_v6\"])\n",
    "    else:\n",
    "        asn6 = None\n",
    "        \n",
    "    if \"country\" in aobj:\n",
    "        cc = aobj['country']\n",
    "    else:\n",
    "        cc = None\n",
    "    \n",
    "    if \"geometry\" in aobj and \"coordinates\" in aobj['geometry']:\n",
    "        (lon, lat) = aobj['geometry']['coordinates']\n",
    "    elif \"longitude\" in aobj and \"latitude\" in aobj:\n",
    "        lon = aobj['longitude']\n",
    "        lat = aobj['latitude']\n",
    "    else:\n",
    "        lon = None\n",
    "        lat = None\n",
    "\n",
    "    return AtlasAnchor(aid, name, pid, ip4, ip6, asn4, asn6, cc, lon, lat)\n",
    "\n",
    "def anchor_dataframe_from_v2api():\n",
    "    data = []\n",
    "    url = \"https://atlas.ripe.net/api/v2/anchors/\"\n",
    "\n",
    "    # iterate over API pagination\n",
    "    while url is not None:\n",
    "        res = requests.get(url)\n",
    "        if not res.ok:\n",
    "            raise RuntimeError(\"Atlas probe API request failed: \"+repr(res.json()))\n",
    "\n",
    "        api_content = json.loads(res.content.decode(\"utf-8\"))\n",
    "        url = api_content['next']\n",
    "        for aobj in api_content[\"results\"]:\n",
    "            data.append(extract_atlas_anchor(aobj))\n",
    "            \n",
    "    # create a dataframe from it\n",
    "    df = pd.DataFrame(data, columns=AtlasAnchor._fields)\n",
    "    \n",
    "    # indexed by probe ID\n",
    "    df.index = df['aid']\n",
    "    del(df['aid'])\n",
    "    \n",
    "    # and return it\n",
    "    return df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "anchor_df = anchor_dataframe_from_v2api()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Anchoring measurement metadata"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "AnchoringMetadata = namedtuple(\"AnchoringMetadata\", (\"aid\", \"msm\", \"af\", \"proto\", \"start\", \"stop\", \"probe_ct\"))\n",
    "\n",
    "def anchoring_measurements_from_v2api(how_many = None):\n",
    "    data = []\n",
    "    url = \"https://atlas.ripe.net/api/v2/anchor-measurements/?include=measurement\"\n",
    "\n",
    "    # iterate over API pagination\n",
    "    while url is not None:\n",
    "        res = requests.get(url)\n",
    "        if not res.ok:\n",
    "            raise RuntimeError(\"Atlas probe API request failed: \"+repr(res.json()))\n",
    "\n",
    "        api_content = json.loads(res.content.decode(\"utf-8\"))\n",
    "        url = api_content['next']\n",
    "        for mobj in api_content[\"results\"]:\n",
    "            try:\n",
    "                aid = int(mobj[\"target\"].strip(\"/\").split(\"/\")[-1])\n",
    "                typ = mobj[\"type\"]\n",
    "                msm = int(mobj[\"measurement\"][\"id\"])\n",
    "                af = int(mobj[\"measurement\"][\"af\"])\n",
    "                if typ == \"traceroute\":\n",
    "                    if \"protocol\" in mobj[\"measurement\"]:\n",
    "                        proto = mobj[\"measurement\"][\"protocol\"] + \"_TR\"\n",
    "                    else:\n",
    "                        proto = \"ICMP_TR\"\n",
    "                elif typ == \"ping\":\n",
    "                    proto = \"ICMP_PING\"\n",
    "                elif typ == \"http\":\n",
    "                    proto = \"TCP_HTTP\"\n",
    "                else:\n",
    "                    proto = None\n",
    "                start = mobj[\"measurement\"][\"start_time\"]\n",
    "                stop = mobj[\"measurement\"][\"stop_time\"]\n",
    "                probe_ct = mobj[\"measurement\"][\"participant_count\"]\n",
    "            except Exception:\n",
    "                continue\n",
    "                \n",
    "            data.append(AnchoringMetadata(aid, msm, af, proto, start, stop, probe_ct))\n",
    "        \n",
    "        if how_many is not None and len(data) >= how_many:\n",
    "            break\n",
    "            \n",
    "    # create a dataframe from it\n",
    "    df = pd.DataFrame(data, columns=AnchoringMetadata._fields)\n",
    "    \n",
    "    # indexed by MSM ID\n",
    "    df.index = df['msm']\n",
    "    del(df['msm'])\n",
    "    \n",
    "    # and return it\n",
    "    return df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "amsm_df = anchoring_measurements_from_v2api()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "with pd.HDFStore(STORE_PATH) as store:\n",
    "    store['amsm_df'] = amsm_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "with pd.HDFStore(STORE_PATH) as store:\n",
    "     amsm_df = store['amsm_df']"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### MSM ID extraction"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Turn anchor names into anchor IDs\n",
    "aid_by_name = anchor_df.loc[:,('name',)]\n",
    "aid_by_name['aid'] = aid_by_name.index\n",
    "aid_by_name.index = aid_by_name['name']\n",
    "del aid_by_name['name']\n",
    "ANCHORS_WE_LIKE = [aid_by_name.loc[aname]['aid'] for aname in ANCHOR_NAMES_WE_LIKE]\n",
    "\n",
    "# turn anchor IDs and protocols into MSMs\n",
    "amsm2_df = amsm_df[amsm_df['aid'].isin(ANCHORS_WE_LIKE)]\n",
    "amsm3_df = amsm2_df[amsm2_df['proto'].isin(PROTOCOLS_WE_LIKE)]\n",
    "MSMS_WE_LIKE = amsm3_df.index.values\n",
    "                    \n",
    "amsm2_df = None\n",
    "amsm3_df = None"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([1026363, 1026365, 1026391, 1026393, 1026399, 1026401, 1042255,\n",
       "       1042257, 1043286, 1043288, 1404333, 1404335, 1423186, 1423188,\n",
       "       1437284, 1437286, 1446417, 1446419, 1583040, 1583042, 1589862,\n",
       "       1589864, 1591156, 1591158, 1664873, 1664875, 1665836, 1665838,\n",
       "       1668851, 1668853, 1768005, 1768007, 1769991, 1769993, 1790232,\n",
       "       1790234, 1804069, 1804071, 1849605, 1849607, 1875035, 1875037,\n",
       "       1898895, 1898897, 1990233, 1990235, 2055768, 2055770, 2096533,\n",
       "       2096536, 2395060, 2395062, 2398550, 2398552, 2417650, 2417652,\n",
       "       2456804, 2456806, 3295749, 3295763, 3315653, 3315656, 3614641,\n",
       "       3614644, 3622418, 3622421, 4464317, 4464320, 6931429, 6931432,\n",
       "       6969364, 6969367, 7861646, 7861649, 7970886, 7970889, 8434915,\n",
       "       8434918, 9180598, 9180601, 9180618, 9180621, 9180638, 9180641,\n",
       "       9180652, 9180655, 9180666, 9180669, 9180704, 9180707, 9180711,\n",
       "       9180714, 9180724, 9180727, 9180739, 9180742, 9180751, 9180754,\n",
       "       9180766, 9180769, 9180792, 9180795, 9180928, 9180931, 9180954,\n",
       "       9180957, 9180989, 9180993, 9181093, 9181096, 9181099, 9181102,\n",
       "       9181174, 9181177, 9181253, 9181256, 9181265, 9181268, 9181275,\n",
       "       9181278, 9181281, 9181284, 9181296, 9181299, 9181316, 9181319,\n",
       "       9181351, 9181354, 9181363, 9181366, 9181371, 9181374, 9181396,\n",
       "       9181399, 9181528, 9181531, 9181616, 9181619, 9181692, 9181695,\n",
       "       9181745, 9181748, 9181777, 9181780, 9181820, 9181823, 9183384,\n",
       "       9183387, 9183523, 9183526, 9183540, 9183543, 9183658, 9183661,\n",
       "       9183786, 9183789, 9238669, 9238679])"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "MSMS_WE_LIKE"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### RTT sample dataframe"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Atlas Latency... something? Raw sample of latency between a source and destination at a given point in time\n",
    "Alp = namedtuple(\"Alp\", (\"msm\",\"time\",\"af\",\"proto\",\"hopcount\",\"pid\",\"sip\",\"dip\",\"rtt\"))\n",
    "\n",
    "RTT_NONE = 0.0\n",
    "\n",
    "def gen_alp(msm_ary):\n",
    "    \"\"\"\n",
    "    Generator function yielding Alp instances for latency samples from an \n",
    "    array containing all results in an MSM downloaded from Atlas.\n",
    "    \"\"\"\n",
    "    for a_res in msm_ary:\n",
    "        ### for ping, extract each latency sample into its own row\n",
    "        if a_res['type'] == 'ping':\n",
    "            if \"rcvd\" in a_res:\n",
    "                for x in a_res[\"result\"]:\n",
    "                    rtt = None\n",
    "                    try: \n",
    "                        rtt = float(x)\n",
    "                    except:\n",
    "                        try:\n",
    "                            rtt = float(x['rtt'])\n",
    "                        except:\n",
    "                            pass\n",
    "                    if rtt:\n",
    "                        yield Alp(a_res['msm_id'],\n",
    "                                  int(a_res['timestamp']), \n",
    "                                  a_res['af'], \n",
    "                                  a_res['proto'] + '_PING', \n",
    "                                  0,\n",
    "                                  a_res['prb_id'], \n",
    "                                  a_res['src_addr'], \n",
    "                                  a_res['dst_addr'], \n",
    "                                  int(rtt * 1000))\n",
    "        \n",
    "        # For traceroute, extract each of the latency samples at the last hop into its own row\n",
    "        # FIXME: hacking this to get last N hops, too \n",
    "        elif a_res['type'] == 'traceroute':\n",
    "            if ('result' in a_res) and ('result' in a_res['result'][-1]):\n",
    "                for h_res in a_res['result'][-1]['result']:\n",
    "                    if ('from' in h_res) and ('rtt' in h_res) and (h_res['from'] == a_res['dst_addr']):\n",
    "                        yield Alp(a_res['msm_id'],\n",
    "                                  int(a_res['timestamp']), \n",
    "                                  a_res['af'], \n",
    "                                  a_res['proto'] + '_TR', \n",
    "                                  a_res['result'][-1]['hop'],\n",
    "                                  a_res['prb_id'], \n",
    "                                  a_res['src_addr'], \n",
    "                                  a_res['dst_addr'], \n",
    "                                  h_res['rtt'])\n",
    "\n",
    "        # For HTTP, return each subresult as a separate RTT sample\n",
    "        elif a_res['type'] == 'http':\n",
    "            for r_res in a_res['result']:\n",
    "                if ('res' in r_res) and (r_res['res'] < 400):\n",
    "                    yield Alp(a_res['timestamp'], r_res['af'], 'TCP_HTTP', \n",
    "                              a_res['prb_id'], r_res['src_addr'], r_res['dst_addr'], r_res['rt'])\n",
    "\n",
    "\n",
    "def alps_for_msm_id(msm_id, cachedir=None, start=None, stop=None):\n",
    "    \"\"\"\n",
    "    Given an MSM ID, fetch results from the cache or from the RIPE Atlas API.\n",
    "    Yield each separate result according to the generation function.\n",
    "    \"\"\"\n",
    "    url = \"https://atlas.ripe.net/api/v2/measurements/%u/results/\" % (msm_id,)\n",
    "\n",
    "    params = {\"format\": \"json\"}\n",
    "    if start is not None and stop is not None:\n",
    "        params[\"start\"] = str(start)\n",
    "        params[\"stop\"] = str(stop)\n",
    "    \n",
    "    if cachedir and os.path.isdir(cachedir):\n",
    "        filepath = os.path.join(cachedir, \"measurement\", \"%u.json\" % (msm_id,))\n",
    "\n",
    "        # download if not present\n",
    "        if not os.path.isfile(filepath):\n",
    "            with open(filepath, mode=\"wb\") as file:\n",
    "                print(\"Cache miss, retrieving \"+url)\n",
    "                res = requests.get(url, params=params)\n",
    "\n",
    "                if not res.ok:\n",
    "                    raise \"Atlas measurement API request failed: \"+repr(res.json())\n",
    "                \n",
    "                file.write(res.content)\n",
    "\n",
    "        # then read from cache\n",
    "        with open(filepath) as stream:\n",
    "            yield from gen_alp(json.loads(stream.read()))\n",
    "\n",
    "    else:\n",
    "        # just read from the net\n",
    "        print(\"Retrieving \"+url+\" uncached\")\n",
    "        res = requests.get(url, params=params)\n",
    "        yield from gen_alp(json.loads(res.content.decode(\"utf-8\")))\n",
    "\n",
    "def alp_df_gen(msm_ids, cachedir=None, start=None, stop=None):\n",
    "    for msm_id in msm_ids:\n",
    "        yield from alps_for_msm_id(msm_id, cachedir=cachedir, start=start, stop=stop)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "!mkdir -p $ATLAS_CACHE_PATH/measurement"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "rtt_df = pd.DataFrame(alp_df_gen(MSMS_WE_LIKE, cachedir=ATLAS_CACHE_PATH, start=START_TIME, stop=STOP_TIME))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now join the anchor ID to the RTT data frame, and clean up for export (anchor ID and time columns)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "aid_by_ip4 = anchor_df.loc[:,('ip4',)]\n",
    "aid_by_ip4['aid'] = aid_by_ip4.index\n",
    "aid_by_ip4.index = aid_by_ip4['ip4']\n",
    "del aid_by_ip4['ip4']\n",
    "\n",
    "aid_by_ip6 = anchor_df.loc[:,('ip6',)]\n",
    "aid_by_ip6['aid'] = aid_by_ip6.index\n",
    "aid_by_ip6.index = aid_by_ip6['ip6']\n",
    "del aid_by_ip6['ip6']\n",
    "\n",
    "rtt_df = pd.concat((rtt_df[rtt_df['af']==4].join(aid_by_ip4, on=\"dip\"), \n",
    "                    rtt_df[rtt_df['af']==6].join(aid_by_ip6, on=\"dip\"))).dropna()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "rtt_df['time'] = pd.to_datetime(rtt_df['time'] * 1e9)\n",
    "rtt_df['aid'] = pd.to_numeric(rtt_df['aid'], downcast='unsigned')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### RTT minimum dataframe"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "min_df = pd.DataFrame({'minrtt': rtt_df.groupby(['aid','pid'])['rtt'].min(),\n",
    "                       'rttstd': rtt_df.groupby(['aid','pid'])['rtt'].std(), \n",
    "                             'minhops': rtt_df.groupby(['aid','pid'])['hopcount'].min(),\n",
    "                             'n':      rtt_df.groupby(['aid','pid'])['rtt'].count()})\n",
    "loc_by_aid = anchor_df.loc[:,('lon','lat')]\n",
    "loc_by_aid.columns = ['alon','alat']\n",
    "loc_by_pid = probe_df.loc[:,('lon','lat')]\n",
    "loc_by_pid.columns = ['plon','plat']\n",
    "min_df = loc_by_pid.join(loc_by_aid.join(min_df, how=\"inner\"), how=\"inner\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Dump dataframes to HDF"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "with pd.HDFStore(STORE_PATH) as store:\n",
    "    store[\"rtt_df\"] = rtt_df\n",
    "    store[\"min_df\"] = min_df\n",
    "    store[\"anchor_df\"] = anchor_df\n",
    "    store[\"probe_df\"] = probe_df"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "## Dataset Description\n",
    "\n",
    "plot heatmap of probes and anchors"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAZwAAADoCAYAAADbhVJaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nKIpSFv8H2Qq3S7Yuy+EAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x2062a59e8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from mpl_toolkits.basemap import Basemap\n",
    "\n",
    "# Plot all anchors and probes on the world map\n",
    "aloc_df = min_df.groupby(level=[0]).first().loc[:,['alon','alat']]\n",
    "ploc_df = min_df.groupby(level=[1]).first().loc[:,['plon','plat']]\n",
    "\n",
    "plt.figure(figsize=(7,5))\n",
    "m = Basemap(projection='robin',lon_0=0,resolution='c')\n",
    "m.drawcoastlines(color=\"darkgrey\")\n",
    "m.fillcontinents(color='white',lake_color='lightgrey')\n",
    "m.drawmapboundary(fill_color='#ddddff')\n",
    "for pid, point in ploc_df.iterrows():\n",
    "    x, y = m(point['plon'], point['plat'])\n",
    "    plt.plot(x,y,\".\",alpha=0.05,color=\"#8899aa\")\n",
    "for aid, point in aloc_df.iterrows():\n",
    "    x, y = m(point['alon'], point['alat'])\n",
    "    plt.plot(x,y,\"r+\")\n",
    "plt.title(\"Probe Density and Anchor Locations\")\n",
    "plt.savefig(os.path.join(PLOT_PATH,\"atlasmap.pdf\"))\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th>plon</th>\n",
       "      <th>plat</th>\n",
       "      <th>alon</th>\n",
       "      <th>alat</th>\n",
       "      <th>minhops</th>\n",
       "      <th>minrtt</th>\n",
       "      <th>n</th>\n",
       "      <th>rttstd</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>aid</th>\n",
       "      <th>pid</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th rowspan=\"30\" valign=\"top\">4</th>\n",
       "      <th>43</th>\n",
       "      <td>4.8595</td>\n",
       "      <td>45.7285</td>\n",
       "      <td>7.660600</td>\n",
       "      <td>45.091300</td>\n",
       "      <td>13</td>\n",
       "      <td>26.455</td>\n",
       "      <td>287</td>\n",
       "      <td>3.074791</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>59</th>\n",
       "      <td>6.7795</td>\n",
       "      <td>49.3315</td>\n",
       "      <td>7.660600</td>\n",
       "      <td>45.091300</td>\n",
       "      <td>14</td>\n",
       "      <td>34.686</td>\n",
       "      <td>288</td>\n",
       "      <td>5.164167</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>60</th>\n",
       "      <td>8.7105</td>\n",
       "      <td>46.2475</td>\n",
       "      <td>7.660600</td>\n",
       "      <td>45.091300</td>\n",
       "      <td>9</td>\n",
       "      <td>19.440</td>\n",
       "      <td>573</td>\n",
       "      <td>3.322747</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>82</th>\n",
       "      <td>9.8115</td>\n",
       "      <td>52.3885</td>\n",
       "      <td>7.660600</td>\n",
       "      <td>45.091300</td>\n",
       "      <td>8</td>\n",
       "      <td>31.956</td>\n",
       "      <td>288</td>\n",
       "      <td>1.203472</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>114</th>\n",
       "      <td>19.1375</td>\n",
       "      <td>48.7395</td>\n",
       "      <td>7.660600</td>\n",
       "      <td>45.091300</td>\n",
       "      <td>13</td>\n",
       "      <td>46.269</td>\n",
       "      <td>287</td>\n",
       "      <td>3.404528</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>142</th>\n",
       "      <td>9.0995</td>\n",
       "      <td>45.4785</td>\n",
       "      <td>7.660600</td>\n",
       "      <td>45.091300</td>\n",
       "      <td>5</td>\n",
       "      <td>4.371</td>\n",
       "      <td>288</td>\n",
       "      <td>0.164339</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>144</th>\n",
       "      <td>13.3195</td>\n",
       "      <td>41.6195</td>\n",
       "      <td>7.660600</td>\n",
       "      <td>45.091300</td>\n",
       "      <td>6</td>\n",
       "      <td>15.500</td>\n",
       "      <td>288</td>\n",
       "      <td>1.225712</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>148</th>\n",
       "      <td>14.4695</td>\n",
       "      <td>50.0815</td>\n",
       "      <td>7.660600</td>\n",
       "      <td>45.091300</td>\n",
       "      <td>6</td>\n",
       "      <td>23.711</td>\n",
       "      <td>288</td>\n",
       "      <td>2.859293</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>241</th>\n",
       "      <td>37.6195</td>\n",
       "      <td>55.7575</td>\n",
       "      <td>7.660600</td>\n",
       "      <td>45.091300</td>\n",
       "      <td>12</td>\n",
       "      <td>49.122</td>\n",
       "      <td>285</td>\n",
       "      <td>4.421772</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>242</th>\n",
       "      <td>18.4895</td>\n",
       "      <td>-33.9815</td>\n",
       "      <td>7.660600</td>\n",
       "      <td>45.091300</td>\n",
       "      <td>11</td>\n",
       "      <td>181.991</td>\n",
       "      <td>288</td>\n",
       "      <td>13.164331</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>249</th>\n",
       "      <td>12.4895</td>\n",
       "      <td>41.8915</td>\n",
       "      <td>7.660600</td>\n",
       "      <td>45.091300</td>\n",
       "      <td>7</td>\n",
       "      <td>13.598</td>\n",
       "      <td>288</td>\n",
       "      <td>5.061152</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>274</th>\n",
       "      <td>11.0805</td>\n",
       "      <td>49.4495</td>\n",
       "      <td>7.660600</td>\n",
       "      <td>45.091300</td>\n",
       "      <td>11</td>\n",
       "      <td>18.219</td>\n",
       "      <td>288</td>\n",
       "      <td>4.073691</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>282</th>\n",
       "      <td>19.0615</td>\n",
       "      <td>47.5685</td>\n",
       "      <td>7.660600</td>\n",
       "      <td>45.091300</td>\n",
       "      <td>12</td>\n",
       "      <td>41.604</td>\n",
       "      <td>288</td>\n",
       "      <td>2.195709</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>331</th>\n",
       "      <td>-117.2315</td>\n",
       "      <td>32.8775</td>\n",
       "      <td>7.660600</td>\n",
       "      <td>45.091300</td>\n",
       "      <td>27</td>\n",
       "      <td>181.265</td>\n",
       "      <td>288</td>\n",
       "      <td>11.384163</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>341</th>\n",
       "      <td>-3.9805</td>\n",
       "      <td>40.5975</td>\n",
       "      <td>7.660600</td>\n",
       "      <td>45.091300</td>\n",
       "      <td>11</td>\n",
       "      <td>36.372</td>\n",
       "      <td>285</td>\n",
       "      <td>4.750160</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>357</th>\n",
       "      <td>16.4095</td>\n",
       "      <td>48.2705</td>\n",
       "      <td>7.660600</td>\n",
       "      <td>45.091300</td>\n",
       "      <td>8</td>\n",
       "      <td>25.969</td>\n",
       "      <td>288</td>\n",
       "      <td>1.380532</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>373</th>\n",
       "      <td>12.2605</td>\n",
       "      <td>57.7685</td>\n",
       "      <td>7.660600</td>\n",
       "      <td>45.091300</td>\n",
       "      <td>14</td>\n",
       "      <td>37.987</td>\n",
       "      <td>288</td>\n",
       "      <td>0.469926</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>376</th>\n",
       "      <td>12.5205</td>\n",
       "      <td>41.9105</td>\n",
       "      <td>7.660600</td>\n",
       "      <td>45.091300</td>\n",
       "      <td>16</td>\n",
       "      <td>15.970</td>\n",
       "      <td>288</td>\n",
       "      <td>2.258522</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>384</th>\n",
       "      <td>10.7715</td>\n",
       "      <td>59.9495</td>\n",
       "      <td>7.660600</td>\n",
       "      <td>45.091300</td>\n",
       "      <td>18</td>\n",
       "      <td>43.985</td>\n",
       "      <td>288</td>\n",
       "      <td>29.136022</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>392</th>\n",
       "      <td>23.8015</td>\n",
       "      <td>38.0485</td>\n",
       "      <td>7.660600</td>\n",
       "      <td>45.091300</td>\n",
       "      <td>10</td>\n",
       "      <td>87.374</td>\n",
       "      <td>285</td>\n",
       "      <td>26.837478</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>407</th>\n",
       "      <td>8.6675</td>\n",
       "      <td>50.1105</td>\n",
       "      <td>7.660600</td>\n",
       "      <td>45.091300</td>\n",
       "      <td>7</td>\n",
       "      <td>14.199</td>\n",
       "      <td>288</td>\n",
       "      <td>2.010609</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>452</th>\n",
       "      <td>7.5795</td>\n",
       "      <td>47.5485</td>\n",
       "      <td>7.660600</td>\n",
       "      <td>45.091300</td>\n",
       "      <td>6</td>\n",
       "      <td>18.445</td>\n",
       "      <td>288</td>\n",
       "      <td>0.560982</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>461</th>\n",
       "      <td>11.6005</td>\n",
       "      <td>48.1875</td>\n",
       "      <td>7.660600</td>\n",
       "      <td>45.091300</td>\n",
       "      <td>9</td>\n",
       "      <td>20.408</td>\n",
       "      <td>288</td>\n",
       "      <td>4.942212</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>567</th>\n",
       "      <td>10.1805</td>\n",
       "      <td>36.8205</td>\n",
       "      <td>7.660600</td>\n",
       "      <td>45.091300</td>\n",
       "      <td>6</td>\n",
       "      <td>35.598</td>\n",
       "      <td>288</td>\n",
       "      <td>11.825796</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>617</th>\n",
       "      <td>12.7075</td>\n",
       "      <td>42.9605</td>\n",
       "      <td>7.660600</td>\n",
       "      <td>45.091300</td>\n",
       "      <td>6</td>\n",
       "      <td>17.890</td>\n",
       "      <td>288</td>\n",
       "      <td>0.400368</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>619</th>\n",
       "      <td>15.4275</td>\n",
       "      <td>46.7515</td>\n",
       "      <td>7.660600</td>\n",
       "      <td>45.091300</td>\n",
       "      <td>8</td>\n",
       "      <td>34.429</td>\n",
       "      <td>288</td>\n",
       "      <td>20.561676</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>626</th>\n",
       "      <td>37.5685</td>\n",
       "      <td>54.1805</td>\n",
       "      <td>7.660600</td>\n",
       "      <td>45.091300</td>\n",
       "      <td>9</td>\n",
       "      <td>55.175</td>\n",
       "      <td>288</td>\n",
       "      <td>1.736979</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>632</th>\n",
       "      <td>-3.9815</td>\n",
       "      <td>40.5985</td>\n",
       "      <td>7.660600</td>\n",
       "      <td>45.091300</td>\n",
       "      <td>11</td>\n",
       "      <td>50.304</td>\n",
       "      <td>179</td>\n",
       "      <td>40.288437</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>668</th>\n",
       "      <td>35.3705</td>\n",
       "      <td>52.3315</td>\n",
       "      <td>7.660600</td>\n",
       "      <td>45.091300</td>\n",
       "      <td>11</td>\n",
       "      <td>64.567</td>\n",
       "      <td>287</td>\n",
       "      <td>3.924490</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>681</th>\n",
       "      <td>33.1115</td>\n",
       "      <td>48.6675</td>\n",
       "      <td>7.660600</td>\n",
       "      <td>45.091300</td>\n",
       "      <td>7</td>\n",
       "      <td>51.057</td>\n",
       "      <td>576</td>\n",
       "      <td>5.730370</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"30\" valign=\"top\">1038</th>\n",
       "      <th>33829</th>\n",
       "      <td>25.0985</td>\n",
       "      <td>60.3815</td>\n",
       "      <td>-1.514821</td>\n",
       "      <td>12.366974</td>\n",
       "      <td>14</td>\n",
       "      <td>181.088</td>\n",
       "      <td>279</td>\n",
       "      <td>136.031021</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33832</th>\n",
       "      <td>3.0385</td>\n",
       "      <td>50.5785</td>\n",
       "      <td>-1.514821</td>\n",
       "      <td>12.366974</td>\n",
       "      <td>14</td>\n",
       "      <td>150.209</td>\n",
       "      <td>276</td>\n",
       "      <td>140.504562</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33835</th>\n",
       "      <td>4.7085</td>\n",
       "      <td>50.8695</td>\n",
       "      <td>-1.514821</td>\n",
       "      <td>12.366974</td>\n",
       "      <td>12</td>\n",
       "      <td>170.712</td>\n",
       "      <td>277</td>\n",
       "      <td>132.986962</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33849</th>\n",
       "      <td>35.1695</td>\n",
       "      <td>31.7915</td>\n",
       "      <td>-1.514821</td>\n",
       "      <td>12.366974</td>\n",
       "      <td>12</td>\n",
       "      <td>204.518</td>\n",
       "      <td>281</td>\n",
       "      <td>139.513968</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33852</th>\n",
       "      <td>8.5395</td>\n",
       "      <td>47.3785</td>\n",
       "      <td>-1.514821</td>\n",
       "      <td>12.366974</td>\n",
       "      <td>15</td>\n",
       "      <td>161.285</td>\n",
       "      <td>282</td>\n",
       "      <td>134.880225</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33853</th>\n",
       "      <td>2.2895</td>\n",
       "      <td>48.6885</td>\n",
       "      <td>-1.514821</td>\n",
       "      <td>12.366974</td>\n",
       "      <td>9</td>\n",
       "      <td>154.978</td>\n",
       "      <td>281</td>\n",
       "      <td>146.874644</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33854</th>\n",
       "      <td>16.3395</td>\n",
       "      <td>48.2295</td>\n",
       "      <td>-1.514821</td>\n",
       "      <td>12.366974</td>\n",
       "      <td>14</td>\n",
       "      <td>219.776</td>\n",
       "      <td>277</td>\n",
       "      <td>135.826283</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33855</th>\n",
       "      <td>30.5095</td>\n",
       "      <td>50.4495</td>\n",
       "      <td>-1.514821</td>\n",
       "      <td>12.366974</td>\n",
       "      <td>11</td>\n",
       "      <td>176.198</td>\n",
       "      <td>280</td>\n",
       "      <td>143.662458</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33857</th>\n",
       "      <td>106.8295</td>\n",
       "      <td>-6.2295</td>\n",
       "      <td>-1.514821</td>\n",
       "      <td>12.366974</td>\n",
       "      <td>10</td>\n",
       "      <td>420.720</td>\n",
       "      <td>280</td>\n",
       "      <td>162.321779</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33859</th>\n",
       "      <td>106.9095</td>\n",
       "      <td>47.8915</td>\n",
       "      <td>-1.514821</td>\n",
       "      <td>12.366974</td>\n",
       "      <td>16</td>\n",
       "      <td>300.018</td>\n",
       "      <td>278</td>\n",
       "      <td>149.568792</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33875</th>\n",
       "      <td>9.2005</td>\n",
       "      <td>48.7895</td>\n",
       "      <td>-1.514821</td>\n",
       "      <td>12.366974</td>\n",
       "      <td>16</td>\n",
       "      <td>155.007</td>\n",
       "      <td>281</td>\n",
       "      <td>177.685542</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33888</th>\n",
       "      <td>1.4515</td>\n",
       "      <td>43.6315</td>\n",
       "      <td>-1.514821</td>\n",
       "      <td>12.366974</td>\n",
       "      <td>11</td>\n",
       "      <td>119.628</td>\n",
       "      <td>282</td>\n",
       "      <td>164.747013</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33896</th>\n",
       "      <td>13.4415</td>\n",
       "      <td>52.6605</td>\n",
       "      <td>-1.514821</td>\n",
       "      <td>12.366974</td>\n",
       "      <td>17</td>\n",
       "      <td>149.739</td>\n",
       "      <td>278</td>\n",
       "      <td>192.145127</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33907</th>\n",
       "      <td>5.7575</td>\n",
       "      <td>49.5205</td>\n",
       "      <td>-1.514821</td>\n",
       "      <td>12.366974</td>\n",
       "      <td>11</td>\n",
       "      <td>181.719</td>\n",
       "      <td>278</td>\n",
       "      <td>147.578539</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33917</th>\n",
       "      <td>-0.6925</td>\n",
       "      <td>45.3905</td>\n",
       "      <td>-1.514821</td>\n",
       "      <td>12.366974</td>\n",
       "      <td>13</td>\n",
       "      <td>151.045</td>\n",
       "      <td>282</td>\n",
       "      <td>164.049276</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33918</th>\n",
       "      <td>8.7075</td>\n",
       "      <td>47.3515</td>\n",
       "      <td>-1.514821</td>\n",
       "      <td>12.366974</td>\n",
       "      <td>14</td>\n",
       "      <td>161.934</td>\n",
       "      <td>280</td>\n",
       "      <td>141.382770</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33921</th>\n",
       "      <td>14.4685</td>\n",
       "      <td>50.1175</td>\n",
       "      <td>-1.514821</td>\n",
       "      <td>12.366974</td>\n",
       "      <td>14</td>\n",
       "      <td>178.421</td>\n",
       "      <td>282</td>\n",
       "      <td>130.870187</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33923</th>\n",
       "      <td>4.8685</td>\n",
       "      <td>45.7485</td>\n",
       "      <td>-1.514821</td>\n",
       "      <td>12.366974</td>\n",
       "      <td>13</td>\n",
       "      <td>178.337</td>\n",
       "      <td>283</td>\n",
       "      <td>147.369607</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33930</th>\n",
       "      <td>-6.0015</td>\n",
       "      <td>37.3975</td>\n",
       "      <td>-1.514821</td>\n",
       "      <td>12.366974</td>\n",
       "      <td>12</td>\n",
       "      <td>215.420</td>\n",
       "      <td>282</td>\n",
       "      <td>130.387186</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33937</th>\n",
       "      <td>6.0785</td>\n",
       "      <td>49.6605</td>\n",
       "      <td>-1.514821</td>\n",
       "      <td>12.366974</td>\n",
       "      <td>10</td>\n",
       "      <td>157.628</td>\n",
       "      <td>281</td>\n",
       "      <td>139.020984</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33938</th>\n",
       "      <td>8.5385</td>\n",
       "      <td>47.3915</td>\n",
       "      <td>-1.514821</td>\n",
       "      <td>12.366974</td>\n",
       "      <td>10</td>\n",
       "      <td>193.502</td>\n",
       "      <td>282</td>\n",
       "      <td>147.883858</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33948</th>\n",
       "      <td>-0.0505</td>\n",
       "      <td>39.9915</td>\n",
       "      <td>-1.514821</td>\n",
       "      <td>12.366974</td>\n",
       "      <td>11</td>\n",
       "      <td>207.139</td>\n",
       "      <td>280</td>\n",
       "      <td>149.713938</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33949</th>\n",
       "      <td>15.6395</td>\n",
       "      <td>60.6115</td>\n",
       "      <td>-1.514821</td>\n",
       "      <td>12.366974</td>\n",
       "      <td>12</td>\n",
       "      <td>207.134</td>\n",
       "      <td>282</td>\n",
       "      <td>152.622385</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33951</th>\n",
       "      <td>22.6695</td>\n",
       "      <td>50.0075</td>\n",
       "      <td>-1.514821</td>\n",
       "      <td>12.366974</td>\n",
       "      <td>10</td>\n",
       "      <td>214.255</td>\n",
       "      <td>279</td>\n",
       "      <td>124.133087</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33953</th>\n",
       "      <td>11.9295</td>\n",
       "      <td>45.3885</td>\n",
       "      <td>-1.514821</td>\n",
       "      <td>12.366974</td>\n",
       "      <td>10</td>\n",
       "      <td>197.131</td>\n",
       "      <td>284</td>\n",
       "      <td>136.941509</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33955</th>\n",
       "      <td>-68.8505</td>\n",
       "      <td>-32.9105</td>\n",
       "      <td>-1.514821</td>\n",
       "      <td>12.366974</td>\n",
       "      <td>13</td>\n",
       "      <td>445.101</td>\n",
       "      <td>15</td>\n",
       "      <td>8.662308</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33972</th>\n",
       "      <td>4.4605</td>\n",
       "      <td>51.9185</td>\n",
       "      <td>-1.514821</td>\n",
       "      <td>12.366974</td>\n",
       "      <td>9</td>\n",
       "      <td>188.668</td>\n",
       "      <td>280</td>\n",
       "      <td>135.068819</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33979</th>\n",
       "      <td>3.0405</td>\n",
       "      <td>50.5815</td>\n",
       "      <td>-1.514821</td>\n",
       "      <td>12.366974</td>\n",
       "      <td>14</td>\n",
       "      <td>150.667</td>\n",
       "      <td>282</td>\n",
       "      <td>130.819539</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33992</th>\n",
       "      <td>7.9715</td>\n",
       "      <td>48.4785</td>\n",
       "      <td>-1.514821</td>\n",
       "      <td>12.366974</td>\n",
       "      <td>9</td>\n",
       "      <td>183.835</td>\n",
       "      <td>284</td>\n",
       "      <td>141.515219</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34017</th>\n",
       "      <td>76.4075</td>\n",
       "      <td>9.9805</td>\n",
       "      <td>-1.514821</td>\n",
       "      <td>12.366974</td>\n",
       "      <td>16</td>\n",
       "      <td>442.453</td>\n",
       "      <td>276</td>\n",
       "      <td>152.770038</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>22120 rows × 8 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                plon     plat      alon       alat  minhops   minrtt    n  \\\n",
       "aid  pid                                                                    \n",
       "4    43       4.8595  45.7285  7.660600  45.091300       13   26.455  287   \n",
       "     59       6.7795  49.3315  7.660600  45.091300       14   34.686  288   \n",
       "     60       8.7105  46.2475  7.660600  45.091300        9   19.440  573   \n",
       "     82       9.8115  52.3885  7.660600  45.091300        8   31.956  288   \n",
       "     114     19.1375  48.7395  7.660600  45.091300       13   46.269  287   \n",
       "     142      9.0995  45.4785  7.660600  45.091300        5    4.371  288   \n",
       "     144     13.3195  41.6195  7.660600  45.091300        6   15.500  288   \n",
       "     148     14.4695  50.0815  7.660600  45.091300        6   23.711  288   \n",
       "     241     37.6195  55.7575  7.660600  45.091300       12   49.122  285   \n",
       "     242     18.4895 -33.9815  7.660600  45.091300       11  181.991  288   \n",
       "     249     12.4895  41.8915  7.660600  45.091300        7   13.598  288   \n",
       "     274     11.0805  49.4495  7.660600  45.091300       11   18.219  288   \n",
       "     282     19.0615  47.5685  7.660600  45.091300       12   41.604  288   \n",
       "     331   -117.2315  32.8775  7.660600  45.091300       27  181.265  288   \n",
       "     341     -3.9805  40.5975  7.660600  45.091300       11   36.372  285   \n",
       "     357     16.4095  48.2705  7.660600  45.091300        8   25.969  288   \n",
       "     373     12.2605  57.7685  7.660600  45.091300       14   37.987  288   \n",
       "     376     12.5205  41.9105  7.660600  45.091300       16   15.970  288   \n",
       "     384     10.7715  59.9495  7.660600  45.091300       18   43.985  288   \n",
       "     392     23.8015  38.0485  7.660600  45.091300       10   87.374  285   \n",
       "     407      8.6675  50.1105  7.660600  45.091300        7   14.199  288   \n",
       "     452      7.5795  47.5485  7.660600  45.091300        6   18.445  288   \n",
       "     461     11.6005  48.1875  7.660600  45.091300        9   20.408  288   \n",
       "     567     10.1805  36.8205  7.660600  45.091300        6   35.598  288   \n",
       "     617     12.7075  42.9605  7.660600  45.091300        6   17.890  288   \n",
       "     619     15.4275  46.7515  7.660600  45.091300        8   34.429  288   \n",
       "     626     37.5685  54.1805  7.660600  45.091300        9   55.175  288   \n",
       "     632     -3.9815  40.5985  7.660600  45.091300       11   50.304  179   \n",
       "     668     35.3705  52.3315  7.660600  45.091300       11   64.567  287   \n",
       "     681     33.1115  48.6675  7.660600  45.091300        7   51.057  576   \n",
       "...              ...      ...       ...        ...      ...      ...  ...   \n",
       "1038 33829   25.0985  60.3815 -1.514821  12.366974       14  181.088  279   \n",
       "     33832    3.0385  50.5785 -1.514821  12.366974       14  150.209  276   \n",
       "     33835    4.7085  50.8695 -1.514821  12.366974       12  170.712  277   \n",
       "     33849   35.1695  31.7915 -1.514821  12.366974       12  204.518  281   \n",
       "     33852    8.5395  47.3785 -1.514821  12.366974       15  161.285  282   \n",
       "     33853    2.2895  48.6885 -1.514821  12.366974        9  154.978  281   \n",
       "     33854   16.3395  48.2295 -1.514821  12.366974       14  219.776  277   \n",
       "     33855   30.5095  50.4495 -1.514821  12.366974       11  176.198  280   \n",
       "     33857  106.8295  -6.2295 -1.514821  12.366974       10  420.720  280   \n",
       "     33859  106.9095  47.8915 -1.514821  12.366974       16  300.018  278   \n",
       "     33875    9.2005  48.7895 -1.514821  12.366974       16  155.007  281   \n",
       "     33888    1.4515  43.6315 -1.514821  12.366974       11  119.628  282   \n",
       "     33896   13.4415  52.6605 -1.514821  12.366974       17  149.739  278   \n",
       "     33907    5.7575  49.5205 -1.514821  12.366974       11  181.719  278   \n",
       "     33917   -0.6925  45.3905 -1.514821  12.366974       13  151.045  282   \n",
       "     33918    8.7075  47.3515 -1.514821  12.366974       14  161.934  280   \n",
       "     33921   14.4685  50.1175 -1.514821  12.366974       14  178.421  282   \n",
       "     33923    4.8685  45.7485 -1.514821  12.366974       13  178.337  283   \n",
       "     33930   -6.0015  37.3975 -1.514821  12.366974       12  215.420  282   \n",
       "     33937    6.0785  49.6605 -1.514821  12.366974       10  157.628  281   \n",
       "     33938    8.5385  47.3915 -1.514821  12.366974       10  193.502  282   \n",
       "     33948   -0.0505  39.9915 -1.514821  12.366974       11  207.139  280   \n",
       "     33949   15.6395  60.6115 -1.514821  12.366974       12  207.134  282   \n",
       "     33951   22.6695  50.0075 -1.514821  12.366974       10  214.255  279   \n",
       "     33953   11.9295  45.3885 -1.514821  12.366974       10  197.131  284   \n",
       "     33955  -68.8505 -32.9105 -1.514821  12.366974       13  445.101   15   \n",
       "     33972    4.4605  51.9185 -1.514821  12.366974        9  188.668  280   \n",
       "     33979    3.0405  50.5815 -1.514821  12.366974       14  150.667  282   \n",
       "     33992    7.9715  48.4785 -1.514821  12.366974        9  183.835  284   \n",
       "     34017   76.4075   9.9805 -1.514821  12.366974       16  442.453  276   \n",
       "\n",
       "                rttstd  \n",
       "aid  pid                \n",
       "4    43       3.074791  \n",
       "     59       5.164167  \n",
       "     60       3.322747  \n",
       "     82       1.203472  \n",
       "     114      3.404528  \n",
       "     142      0.164339  \n",
       "     144      1.225712  \n",
       "     148      2.859293  \n",
       "     241      4.421772  \n",
       "     242     13.164331  \n",
       "     249      5.061152  \n",
       "     274      4.073691  \n",
       "     282      2.195709  \n",
       "     331     11.384163  \n",
       "     341      4.750160  \n",
       "     357      1.380532  \n",
       "     373      0.469926  \n",
       "     376      2.258522  \n",
       "     384     29.136022  \n",
       "     392     26.837478  \n",
       "     407      2.010609  \n",
       "     452      0.560982  \n",
       "     461      4.942212  \n",
       "     567     11.825796  \n",
       "     617      0.400368  \n",
       "     619     20.561676  \n",
       "     626      1.736979  \n",
       "     632     40.288437  \n",
       "     668      3.924490  \n",
       "     681      5.730370  \n",
       "...                ...  \n",
       "1038 33829  136.031021  \n",
       "     33832  140.504562  \n",
       "     33835  132.986962  \n",
       "     33849  139.513968  \n",
       "     33852  134.880225  \n",
       "     33853  146.874644  \n",
       "     33854  135.826283  \n",
       "     33855  143.662458  \n",
       "     33857  162.321779  \n",
       "     33859  149.568792  \n",
       "     33875  177.685542  \n",
       "     33888  164.747013  \n",
       "     33896  192.145127  \n",
       "     33907  147.578539  \n",
       "     33917  164.049276  \n",
       "     33918  141.382770  \n",
       "     33921  130.870187  \n",
       "     33923  147.369607  \n",
       "     33930  130.387186  \n",
       "     33937  139.020984  \n",
       "     33938  147.883858  \n",
       "     33948  149.713938  \n",
       "     33949  152.622385  \n",
       "     33951  124.133087  \n",
       "     33953  136.941509  \n",
       "     33955    8.662308  \n",
       "     33972  135.068819  \n",
       "     33979  130.819539  \n",
       "     33992  141.515219  \n",
       "     34017  152.770038  \n",
       "\n",
       "[22120 rows x 8 columns]"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "min_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}