{
  "id": 287299,
  "title": "Why are some VWAP values inf?",
  "url": "/competitions/g-research-crypto-forecasting/discussion/287299",
  "author_name": "",
  "post_date": "2021-11-13T05:18:25.735565500Z",
  "votes": 8,
  "comment_count": 1,
  "views": 0,
  "content": "<pre><code>$ grep inf train.csv \n1592193180,10,51.0,477.39785,480.91,471.9999,476.9179,61.64613999999999,inf,0.0042344692129800965\n1592251080,10,12.0,522.36745,526.6249,518.1,522.0478499999998,2.320164000000001,inf,0.002794404324555355\n1592333460,10,25.0,530.2927,530.9067,529.87,529.9846,7.682914999999999,inf,0.014912622339746884\n1592560440,10,7.0,515.25575,517.1,513.4096,516.0128,0.197087,inf,\n1592680740,10,5.0,512.0046,512.9992,511.0,512.0046,0.5,inf,0.0003856143575124982\n1592824560,10,6.0,503.6,503.6,502.0,503.6,0.0,-inf,-0.002349268400027915\n1592825580,10,20.0,500.9,502.0,499.8,500.97995,8.106645,inf,0.004401760704281665\n</code></pre>",
  "messages": [
    {
      "id": "1580763",
      "postDate": "11/13/2021 05:18:25",
      "content": "<pre><code>$ grep inf train.csv \n1592193180,10,51.0,477.39785,480.91,471.9999,476.9179,61.64613999999999,inf,0.0042344692129800965\n1592251080,10,12.0,522.36745,526.6249,518.1,522.0478499999998,2.320164000000001,inf,0.002794404324555355\n1592333460,10,25.0,530.2927,530.9067,529.87,529.9846,7.682914999999999,inf,0.014912622339746884\n1592560440,10,7.0,515.25575,517.1,513.4096,516.0128,0.197087,inf,\n1592680740,10,5.0,512.0046,512.9992,511.0,512.0046,0.5,inf,0.0003856143575124982\n1592824560,10,6.0,503.6,503.6,502.0,503.6,0.0,-inf,-0.002349268400027915\n1592825580,10,20.0,500.9,502.0,499.8,500.97995,8.106645,inf,0.004401760704281665\n</code></pre>",
      "rawMarkdown": "```\n$ grep inf train.csv \n1592193180,10,51.0,477.39785,480.91,471.9999,476.9179,61.64613999999999,inf,0.0042344692129800965\n1592251080,10,12.0,522.36745,526.6249,518.1,522.0478499999998,2.320164000000001,inf,0.002794404324555355\n1592333460,10,25.0,530.2927,530.9067,529.87,529.9846,7.682914999999999,inf,0.014912622339746884\n1592560440,10,7.0,515.25575,517.1,513.4096,516.0128,0.197087,inf,\n1592680740,10,5.0,512.0046,512.9992,511.0,512.0046,0.5,inf,0.0003856143575124982\n1592824560,10,6.0,503.6,503.6,502.0,503.6,0.0,-inf,-0.002349268400027915\n1592825580,10,20.0,500.9,502.0,499.8,500.97995,8.106645,inf,0.004401760704281665\n```",
      "votes": null
    },
    {
      "id": "1598859",
      "postDate": "11/28/2021 23:16:54",
      "content": "<p>because some other feature is inf or Nan</p>\n<p>`<br>\nfrom datetime import datetime, timedelta</p>\n<p>date_today = datetime.now()<br>\ndays = pd.date_range(date_today, date_today + timedelta(4), freq='D')</p>\n<h1>initialize list of lists</h1>\n<p>test_data = [<br>\n    100,<br>\n    105, <br>\n    50,<br>\n    np.inf,<br>\n    10,<br>\n]</p>\n<h1>Create the pandas DataFrame</h1>\n<h1>test_df = pd.DataFrame(test_data, columns = ['Volume', 'Price'])</h1>\n<p>test_df = pd.DataFrame({'time': days, 'Volume': test_data, 'Price':[10,15,10,100,10]} )<br>\ntest_df = test_df.set_index('time')<br>\ndf = df.assign(<br>\n    vwap=df.eval(<br>\n        'wgtd = Price * Volume', inplace=False<br>\n    ).groupby(df.index.date).cumsum().eval('wgtd / Volume')<br>\n)<br>\ndf<br>\n`</p>",
      "rawMarkdown": "because some other feature is inf or Nan\n\n`\nfrom datetime import datetime, timedelta\n\ndate_today = datetime.now()\ndays = pd.date_range(date_today, date_today + timedelta(4), freq='D')\n# initialize list of lists\ntest_data = [\n    100,\n    105, \n    50,\n    np.inf,\n    10,\n]\n  \n# Create the pandas DataFrame\n#test_df = pd.DataFrame(test_data, columns = ['Volume', 'Price'])\ntest_df = pd.DataFrame({'time': days, 'Volume': test_data, 'Price':[10,15,10,100,10]} )\ntest_df = test_df.set_index('time')\ndf = df.assign(\n    vwap=df.eval(\n        'wgtd = Price * Volume', inplace=False\n    ).groupby(df.index.date).cumsum().eval('wgtd / Volume')\n)\ndf\n`",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 1598859,
      "author_name": "vladimircape",
      "author_url": "",
      "post_date": "11/28/2021 23:16:54",
      "content": "<p>because some other feature is inf or Nan</p>\n<p>`<br>\nfrom datetime import datetime, timedelta</p>\n<p>date_today = datetime.now()<br>\ndays = pd.date_range(date_today, date_today + timedelta(4), freq='D')</p>\n<h1>initialize list of lists</h1>\n<p>test_data = [<br>\n    100,<br>\n    105, <br>\n    50,<br>\n    np.inf,<br>\n    10,<br>\n]</p>\n<h1>Create the pandas DataFrame</h1>\n<h1>test_df = pd.DataFrame(test_data, columns = ['Volume', 'Price'])</h1>\n<p>test_df = pd.DataFrame({'time': days, 'Volume': test_data, 'Price':[10,15,10,100,10]} )<br>\ntest_df = test_df.set_index('time')<br>\ndf = df.assign(<br>\n    vwap=df.eval(<br>\n        'wgtd = Price * Volume', inplace=False<br>\n    ).groupby(df.index.date).cumsum().eval('wgtd / Volume')<br>\n)<br>\ndf<br>\n`</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "1580763": "```\n$ grep inf train.csv \n1592193180,10,51.0,477.39785,480.91,471.9999,476.9179,61.64613999999999,inf,0.0042344692129800965\n1592251080,10,12.0,522.36745,526.6249,518.1,522.0478499999998,2.320164000000001,inf,0.002794404324555355\n1592333460,10,25.0,530.2927,530.9067,529.87,529.9846,7.682914999999999,inf,0.014912622339746884\n1592560440,10,7.0,515.25575,517.1,513.4096,516.0128,0.197087,inf,\n1592680740,10,5.0,512.0046,512.9992,511.0,512.0046,0.5,inf,0.0003856143575124982\n1592824560,10,6.0,503.6,503.6,502.0,503.6,0.0,-inf,-0.002349268400027915\n1592825580,10,20.0,500.9,502.0,499.8,500.97995,8.106645,inf,0.004401760704281665\n```",
    "1598859": "because some other feature is inf or Nan\n\n`\nfrom datetime import datetime, timedelta\n\ndate_today = datetime.now()\ndays = pd.date_range(date_today, date_today + timedelta(4), freq='D')\n# initialize list of lists\ntest_data = [\n    100,\n    105, \n    50,\n    np.inf,\n    10,\n]\n  \n# Create the pandas DataFrame\n#test_df = pd.DataFrame(test_data, columns = ['Volume', 'Price'])\ntest_df = pd.DataFrame({'time': days, 'Volume': test_data, 'Price':[10,15,10,100,10]} )\ntest_df = test_df.set_index('time')\ndf = df.assign(\n    vwap=df.eval(\n        'wgtd = Price * Volume', inplace=False\n    ).groupby(df.index.date).cumsum().eval('wgtd / Volume')\n)\ndf\n`"
  },
  "source": "meta"
}