Nielsen2014Python_virkdata.ipynb

{
 "metadata": {
  "name": "",
  "signature": "sha256:fe2e9b256bf43cc51353e9b5c421d862646da62e5ff23279534976ad5686a7ef"
 },
 "nbformat": 3,
 "nbformat_minor": 0,
 "worksheets": [
  {
   "cells": [
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "Analysis of Danish financial statements\n",
      "===================================\n",
      "\n",
      "- Data from 'Erhvervsstyrelsen'.\n",
      "- http://datahub.virk.dk/dataset/regnskabsdata-fra-selskaber-sample\n",
      "- Sample of some hundreds of companies\n",
      "- Data in zipped XBRL XML format\n",
      "- http://datahub.virk.dk/sites/default/files/storage/1000_digitale_aarsrapporter.zip"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "from lxml import etree\n",
      "import os\n",
      "import pandas as pd"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 1
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "# The data was unzipped in this directory:\n",
      "dirname = '/home/fnielsen/data/virkdk/1000_digitale_aarsrapporter'"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 2
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "# Attempt on a dirty reading of the data\n",
      "# Not necessarily pretty, - including some strange workarounds\n",
      "rows= []\n",
      "for filename in os.listdir(dirname):\n",
      "    if not filename.endswith('.xml'):\n",
      "        continue\n",
      "    try:\n",
      "        tree = etree.parse(open(os.path.join(dirname, filename)))\n",
      "    except: \n",
      "        # There is XML error in the XML because of spaces!?\n",
      "        continue\n",
      "    d = {}\n",
      "    for element in tree.findall('/'):\n",
      "        if not isinstance(element, etree._Comment) and '{http://xbrl.dcca.dk/' in element.tag:\n",
      "            tag = element.tag[element.tag.index('}') + 1:]\n",
      "            context = element.attrib['contextRef']\n",
      "            if element.text is None:\n",
      "                # For some reason TypeError is not caught!?\n",
      "                value = None\n",
      "            else:\n",
      "                try: \n",
      "                    value = float(element.text)\n",
      "                except TypeError:\n",
      "                    value = element.text\n",
      "                except ValueError:\n",
      "                    value = element.text\n",
      "            d[tag] = value\n",
      "    rows.append(d)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 3
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "df = pd.DataFrame(rows)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 4
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "df['NameAndSurnameOfChairmanOfGeneralMeeting'].value_counts()[:12]"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "metadata": {},
       "output_type": "pyout",
       "prompt_number": 5,
       "text": [
        "Nils B. Bonde             7\n",
        "Michael Vinther           7\n",
        "Jens Tange M\u00f8llmann       5\n",
        "Kim Larsen                5\n",
        "Henrik Klougart           4\n",
        "Uffe Martin Jensen        4\n",
        "Henrik Rasmussen          4\n",
        "Merete Lundbye M\u00f8ller     4\n",
        "Johannes Nielsen          4\n",
        "Poul-Erik Vind            4\n",
        "S\u00f8ren Bruun Rasmussen     3\n",
        "Thomas Folmann            3\n",
        "dtype: int64"
       ]
      }
     ],
     "prompt_number": 5
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "df['NameOfAuditFirm'].value_counts()[:12]"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "metadata": {},
       "output_type": "pyout",
       "prompt_number": 6,
       "text": [
        "Deloitte Statsautoriseret Revisionspartnerselskab               75\n",
        "PricewaterhouseCoopers Statsautoriseret Revisionspartnerselskab    63\n",
        "Beierholm                                                       52\n",
        "BDO Statsautoriseret revisionsaktieselskab                      52\n",
        "KPMG                                                            26\n",
        "AP | Statsautoriserede Revisorer P/S,                           17\n",
        "Redmark, Statsautoriseret Revisionspartnerselskab               15\n",
        "KPMG Statsautoriseret Revisionspartnerselskab                   13\n",
        "GLB REVISION, Statsautoriserede Revisorer A/S                   12\n",
        "CHRISTENSEN KJ\u00c6RULFF, statsautoriseret revisionsaktieselskab    12\n",
        "Partner Revision, statsautoriseret revisionsaktieselskab        11\n",
        "                                                                 8\n",
        "dtype: int64"
       ]
      }
     ],
     "prompt_number": 6
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 7
    }
   ],
   "metadata": {}
  }
 ]
}