Max validation errors

Max validation errors.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Unable to parse whitelist (/home/hadoop/anaconda2/lib/python2.7/site-packages/moztelemetry/histogram-whitelists.json). Assuming all histograms are acceptable.\n"
     ]
    }
   ],
   "source": [
    "from moztelemetry import get_pings, get_pings_properties\n",
    "\n",
    "# filter the ones we care about early...\n",
    "def filterOurs(ping):\n",
    "    try:\n",
    "        syncs = ping[\"payload\"][\"syncs\"]\n",
    "    except KeyError:\n",
    "        return False\n",
    "    for sync in syncs:\n",
    "        for engine in sync.get(\"engines\", []):\n",
    "            if \"name\" in engine and engine[\"name\"] == \"bookmarks\" and \"validation\" in engine and \"problems\" in engine[\"validation\"]:\n",
    "                return True\n",
    "    return False\n",
    "        \n",
    "pings = get_pings(sc, doc_type='sync', fraction=1.0).filter(filterOurs)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{u'clientCycles': 7,\n",
       " u'clientMissing': 84,\n",
       " u'deletedChildren': 14,\n",
       " u'deletedParents': 17,\n",
       " u'differences': 63,\n",
       " u'duplicateChildren': 1,\n",
       " u'missingChildren': 914,\n",
       " u'multipleParents': 17,\n",
       " u'orphans': 62,\n",
       " u'parentChildMismatches': 53,\n",
       " u'parentNotFolder': 17,\n",
       " u'rootOnServer': 1,\n",
       " u'sdiff:childGUIDs': 7,\n",
       " u'sdiff:parentid': 15,\n",
       " u'serverMissing': 62473,\n",
       " u'serverUnexpected': 48,\n",
       " u'structuralDifferences': 22,\n",
       " u'wrongParentName': 62}"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def flattenBookmarkValidations(ping):\n",
    "    result = []\n",
    "    for sync in ping[\"payload\"][\"syncs\"]:\n",
    "        for engine in sync.get(\"engines\", []):\n",
    "            if \"validation\" in engine:\n",
    "                # turn it into a dict.\n",
    "                this = {}\n",
    "                for problem in engine[\"validation\"][\"problems\"]:\n",
    "                    this[problem[\"name\"]] = problem[\"count\"]\n",
    "                result.append(this)\n",
    "    return result\n",
    "\n",
    "# And reduce each one into the max\n",
    "def reduceToMax(a, b):\n",
    "    allKeys = set(a.keys())\n",
    "    allKeys.update(b.keys())\n",
    "    result = {}\n",
    "    for key in allKeys:\n",
    "        result[key] = max(a.get(key, 0), b.get(key, 0))\n",
    "    return result\n",
    "\n",
    "\n",
    "pings.flatMap(flattenBookmarkValidations).reduce(reduceToMax)"
   ]
  }
 ],
 "metadata": {
  "anaconda-cloud": {},
  "kernelspec": {
   "display_name": "Python [default]",
   "language": "python",
   "name": "python2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}

Max validation errors.py

# coding: utf-8

# In[1]:

from moztelemetry import get_pings, get_pings_properties

# filter the ones we care about early...
def filterOurs(ping):
    try:
        syncs = ping["payload"]["syncs"]
    except KeyError:
        return False
    for sync in syncs:
        for engine in sync.get("engines", []):
            if "name" in engine and engine["name"] == "bookmarks" and "validation" in engine and "problems" in engine["validation"]:
                return True
    return False
        
pings = get_pings(sc, doc_type='sync', fraction=1.0).filter(filterOurs)


# In[2]:

def flattenBookmarkValidations(ping):
    result = []
    for sync in ping["payload"]["syncs"]:
        for engine in sync.get("engines", []):
            if "validation" in engine:
                # turn it into a dict.
                this = {}
                for problem in engine["validation"]["problems"]:
                    this[problem["name"]] = problem["count"]
                result.append(this)
    return result

# And reduce each one into the max
def reduceToMax(a, b):
    allKeys = set(a.keys())
    allKeys.update(b.keys())
    result = {}
    for key in allKeys:
        result[key] = max(a.get(key, 0), b.get(key, 0))
    return result


pings.flatMap(flattenBookmarkValidations).reduce(reduceToMax)