{
  "id": 330714,
  "title": "No Matching Signature Found Error When Using ffill()",
  "url": "/competitions/amex-default-prediction/discussion/330714",
  "author_name": "",
  "post_date": "2022-06-13T23:01:02.149796900Z",
  "votes": 1,
  "comment_count": 5,
  "views": 0,
  "content": "<p>Hi Everyone,</p>\n<p>I have been trying to fill the null values of this dataset by using techniques such as pandas ffill() function. However, whenever I try and execute the code df.ffill() I keep receiving an type error saying \"No Matching Signature Found\". I'm having trouble figuring out why this error keeps happening and was wondering if anyone may be able to explain why. </p>",
  "messages": [
    {
      "id": "1819560",
      "postDate": "06/13/2022 23:01:02",
      "content": "<p>Hi Everyone,</p>\n<p>I have been trying to fill the null values of this dataset by using techniques such as pandas ffill() function. However, whenever I try and execute the code df.ffill() I keep receiving an type error saying \"No Matching Signature Found\". I'm having trouble figuring out why this error keeps happening and was wondering if anyone may be able to explain why. </p>",
      "rawMarkdown": "Hi Everyone,\n\nI have been trying to fill the null values of this dataset by using techniques such as pandas ffill() function. However, whenever I try and execute the code df.ffill() I keep receiving an type error saying \"No Matching Signature Found\". I'm having trouble figuring out why this error keeps happening and was wondering if anyone may be able to explain why.",
      "votes": null
    },
    {
      "id": "1819900",
      "postDate": "06/14/2022 08:14:33",
      "content": "<p>what is the version of pandas you are using?</p>",
      "rawMarkdown": "what is the version of pandas you are using?",
      "votes": null
    },
    {
      "id": "1820042",
      "postDate": "06/14/2022 10:22:54",
      "content": "<p>For some records and features values are null for all 13 time steps, therefore no value to ffill with. Maybe this is the issue.</p>",
      "rawMarkdown": "For some records and features values are null for all 13 time steps, therefore no value to ffill with. Maybe this is the issue.",
      "votes": null
    },
    {
      "id": "1820693",
      "postDate": "06/14/2022 21:14:32",
      "content": "<p>I think even if a series contains all null values the ffill function will just fill with all null values. I made a series full of just null values and tested this and that is what happened. I appreciate the response though!</p>",
      "rawMarkdown": "I think even if a series contains all null values the ffill function will just fill with all null values. I made a series full of just null values and tested this and that is what happened. I appreciate the response though!",
      "votes": null
    },
    {
      "id": "1820694",
      "postDate": "06/14/2022 21:16:04",
      "content": "<p>Version 1.3.5</p>",
      "rawMarkdown": "Version 1.3.5",
      "votes": null
    },
    {
      "id": "1822934",
      "postDate": "06/16/2022 20:58:03",
      "content": "<p>Alright I found out why it doesn't work. For some reason the ffill() function does not work for the float16 data type. Once I converted to float64 the function worked.</p>",
      "rawMarkdown": "Alright I found out why it doesn't work. For some reason the ffill() function does not work for the float16 data type. Once I converted to float64 the function worked.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 1819900,
      "author_name": "raddar",
      "author_url": "",
      "post_date": "06/14/2022 08:14:33",
      "content": "<p>what is the version of pandas you are using?</p>",
      "votes": null,
      "replies": [
        {
          "id": 1820694,
          "author_name": "robertturro",
          "author_url": "",
          "post_date": "06/14/2022 21:16:04",
          "content": "<p>Version 1.3.5</p>",
          "votes": null,
          "replies": []
        }
      ]
    },
    {
      "id": 1820042,
      "author_name": "bacicnikola",
      "author_url": "",
      "post_date": "06/14/2022 10:22:54",
      "content": "<p>For some records and features values are null for all 13 time steps, therefore no value to ffill with. Maybe this is the issue.</p>",
      "votes": null,
      "replies": [
        {
          "id": 1820693,
          "author_name": "robertturro",
          "author_url": "",
          "post_date": "06/14/2022 21:14:32",
          "content": "<p>I think even if a series contains all null values the ffill function will just fill with all null values. I made a series full of just null values and tested this and that is what happened. I appreciate the response though!</p>",
          "votes": null,
          "replies": []
        }
      ]
    },
    {
      "id": 1822934,
      "author_name": "robertturro",
      "author_url": "",
      "post_date": "06/16/2022 20:58:03",
      "content": "<p>Alright I found out why it doesn't work. For some reason the ffill() function does not work for the float16 data type. Once I converted to float64 the function worked.</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "1819560": "Hi Everyone,\n\nI have been trying to fill the null values of this dataset by using techniques such as pandas ffill() function. However, whenever I try and execute the code df.ffill() I keep receiving an type error saying \"No Matching Signature Found\". I'm having trouble figuring out why this error keeps happening and was wondering if anyone may be able to explain why.",
    "1819900": "what is the version of pandas you are using?",
    "1820042": "For some records and features values are null for all 13 time steps, therefore no value to ffill with. Maybe this is the issue.",
    "1820693": "I think even if a series contains all null values the ffill function will just fill with all null values. I made a series full of just null values and tested this and that is what happened. I appreciate the response though!",
    "1820694": "Version 1.3.5",
    "1822934": "Alright I found out why it doesn't work. For some reason the ffill() function does not work for the float16 data type. Once I converted to float64 the function worked."
  },
  "source": "meta"
}