Has someone managed to query the train and test set directly from BigQuery in kernel? Would be more nice to do it in this way from a business point of view.
", "messages": [ { "id": "389963", "postDate": "09/19/2018 14:17:21", "content": "Has someone managed to query the train and test set directly from BigQuery in kernel? Would be more nice to do it in this way from a business point of view.
", "rawMarkdown": "Has someone managed to query the train and test set directly from BigQuery in kernel? Would be more nice to do it in this way from a business point of view.", "votes": null }, { "id": "389981", "postDate": "09/19/2018 14:49:03", "content": "Yes, I managed to query data directly from BigQuery. There's no difference with public csv data except for NaN columns.
", "rawMarkdown": "Yes, I managed to query data directly from BigQuery. There's no difference with public csv data except for NaN columns.", "votes": null }, { "id": "389985", "postDate": "09/19/2018 14:56:09", "content": "Okay thanks, although I'm still more interested in querying the data directly from the source (BigQuery) instead of using the csv files as an intermediary step :-)
", "rawMarkdown": "Okay thanks, although I'm still more interested in querying the data directly from the source (BigQuery) instead of using the csv files as an intermediary step :-)", "votes": null }, { "id": "389987", "postDate": "09/19/2018 15:03:19", "content": "Have you already seen the GA BigQuery Starter Kernel ?
", "rawMarkdown": "Have you already seen the [GA BigQuery Starter Kernel][1] ?\n\n\n [1]: https://www.kaggle.com/vanausloos/ga-bigquery-starter-kernel", "votes": null }, { "id": "389989", "postDate": "09/19/2018 15:07:34", "content": "Just what I needed. Thank you very much!
", "rawMarkdown": "Just what I needed. Thank you very much!", "votes": null }, { "id": "390033", "postDate": "09/19/2018 15:46:46", "content": "Hey Stanislav Keavchuk !\nHow do you know that ? Had you gotten useful info from the BigQuery source which presents NaN on public csv? \nI used to play BigQuery for a while ( not this dataset). But writing query is less efficient than using python, any suggestions?
", "rawMarkdown": "Hey Stanislav Keavchuk !\nHow do you know that ? Had you gotten useful info from the BigQuery source which presents NaN on public csv? \nI used to play BigQuery for a while ( not this dataset). But writing query is less efficient than using python, any suggestions?", "votes": null }, { "id": "390038", "postDate": "09/19/2018 15:55:11", "content": "The BigQuery data is exactly the same as CSV data. I compared them char by char. BigQuery data contains some extra columns, but they contain only NaNs.
", "rawMarkdown": "The BigQuery data is exactly the same as CSV data. I compared them char by char. BigQuery data contains some extra columns, but they contain only NaNs.", "votes": null }, { "id": "390039", "postDate": "09/19/2018 15:59:22", "content": "So, the public csv is just a subset of bigquery dataset. But other columns in bigquery is just NaN. Is it ?
", "rawMarkdown": "So, the public csv is just a subset of bigquery dataset. But other columns in bigquery is just NaN. Is it ?", "votes": null }, { "id": "390042", "postDate": "09/19/2018 16:01:24", "content": "It seems like that :-)
", "rawMarkdown": "It seems like that :-)", "votes": null }, { "id": "425674", "postDate": "11/21/2018 23:55:37", "content": "I am wondering how you were able to upload the dataset into google cloud storage, the training set is 1GB, when I unzip it, it becomes 25GB, whether i try to upload it to google cloud storage (to later import the data into my BigQuery Table) from the browser or from the google terminal sdk, it fails in both cases, I am wondering if you had a similar challenge? how were you able to upload the data to google cloud?
", "rawMarkdown": "I am wondering how you were able to upload the dataset into google cloud storage, the training set is 1GB, when I unzip it, it becomes 25GB, whether i try to upload it to google cloud storage (to later import the data into my BigQuery Table) from the browser or from the google terminal sdk, it fails in both cases, I am wondering if you had a similar challenge? how were you able to upload the data to google cloud?", "votes": null } ], "comments": [ { "id": 389981, "author_name": "stantrc", "author_url": "", "post_date": "09/19/2018 14:49:03", "content": "Yes, I managed to query data directly from BigQuery. There's no difference with public csv data except for NaN columns.
", "votes": null, "replies": [ { "id": 389985, "author_name": "madshelt", "author_url": "", "post_date": "09/19/2018 14:56:09", "content": "Okay thanks, although I'm still more interested in querying the data directly from the source (BigQuery) instead of using the csv files as an intermediary step :-)
", "votes": null, "replies": [] }, { "id": 389987, "author_name": "stantrc", "author_url": "", "post_date": "09/19/2018 15:03:19", "content": "Have you already seen the GA BigQuery Starter Kernel ?
", "votes": null, "replies": [] }, { "id": 389989, "author_name": "madshelt", "author_url": "", "post_date": "09/19/2018 15:07:34", "content": "Just what I needed. Thank you very much!
", "votes": null, "replies": [] } ] }, { "id": 390033, "author_name": "ylt0609", "author_url": "", "post_date": "09/19/2018 15:46:46", "content": "Hey Stanislav Keavchuk !\nHow do you know that ? Had you gotten useful info from the BigQuery source which presents NaN on public csv? \nI used to play BigQuery for a while ( not this dataset). But writing query is less efficient than using python, any suggestions?
", "votes": null, "replies": [ { "id": 390038, "author_name": "stantrc", "author_url": "", "post_date": "09/19/2018 15:55:11", "content": "The BigQuery data is exactly the same as CSV data. I compared them char by char. BigQuery data contains some extra columns, but they contain only NaNs.
", "votes": null, "replies": [] }, { "id": 390039, "author_name": "ylt0609", "author_url": "", "post_date": "09/19/2018 15:59:22", "content": "So, the public csv is just a subset of bigquery dataset. But other columns in bigquery is just NaN. Is it ?
", "votes": null, "replies": [] }, { "id": 390042, "author_name": "stantrc", "author_url": "", "post_date": "09/19/2018 16:01:24", "content": "It seems like that :-)
", "votes": null, "replies": [] } ] }, { "id": 425674, "author_name": "hadigol", "author_url": "", "post_date": "11/21/2018 23:55:37", "content": "I am wondering how you were able to upload the dataset into google cloud storage, the training set is 1GB, when I unzip it, it becomes 25GB, whether i try to upload it to google cloud storage (to later import the data into my BigQuery Table) from the browser or from the google terminal sdk, it fails in both cases, I am wondering if you had a similar challenge? how were you able to upload the data to google cloud?
", "votes": null, "replies": [] } ], "raw_markdown_by_id": { "389963": "Has someone managed to query the train and test set directly from BigQuery in kernel? Would be more nice to do it in this way from a business point of view.", "389981": "Yes, I managed to query data directly from BigQuery. There's no difference with public csv data except for NaN columns.", "389985": "Okay thanks, although I'm still more interested in querying the data directly from the source (BigQuery) instead of using the csv files as an intermediary step :-)", "389987": "Have you already seen the [GA BigQuery Starter Kernel][1] ?\n\n\n [1]: https://www.kaggle.com/vanausloos/ga-bigquery-starter-kernel", "389989": "Just what I needed. Thank you very much!", "390033": "Hey Stanislav Keavchuk !\nHow do you know that ? Had you gotten useful info from the BigQuery source which presents NaN on public csv? \nI used to play BigQuery for a while ( not this dataset). But writing query is less efficient than using python, any suggestions?", "390038": "The BigQuery data is exactly the same as CSV data. I compared them char by char. BigQuery data contains some extra columns, but they contain only NaNs.", "390039": "So, the public csv is just a subset of bigquery dataset. But other columns in bigquery is just NaN. Is it ?", "390042": "It seems like that :-)", "425674": "I am wondering how you were able to upload the dataset into google cloud storage, the training set is 1GB, when I unzip it, it becomes 25GB, whether i try to upload it to google cloud storage (to later import the data into my BigQuery Table) from the browser or from the google terminal sdk, it fails in both cases, I am wondering if you had a similar challenge? how were you able to upload the data to google cloud?" }, "source": "meta" }