Some quick tips for getting started with this competition:
\n\nFound the \"ObsPy\" package which can handle seismology data - link\nSpecifically the ObsPy tutorial can be helpful to derive feature - link \nAnother Python package \"pyrocko\" for handling and understanding seismology data - link\nThere are a couple of recent papers published by Los Alamos National Laboratory paper 1 and paper 2
\n\nHope this is useful !!
", "messages": [ { "id": "453999", "postDate": "01/11/2019 02:37:57", "content": "Some quick tips for getting started with this competition:
\n\nFound the \"ObsPy\" package which can handle seismology data - link\nSpecifically the ObsPy tutorial can be helpful to derive feature - link \nAnother Python package \"pyrocko\" for handling and understanding seismology data - link\nThere are a couple of recent papers published by Los Alamos National Laboratory paper 1 and paper 2
\n\nHope this is useful !!
", "rawMarkdown": "Some quick tips for getting started with this competition:\n\nFound the \"ObsPy\" package which can handle seismology data - [link][1]\nSpecifically the ObsPy tutorial can be helpful to derive feature - [link][2] \nAnother Python package \"pyrocko\" for handling and understanding seismology data - [link][3]\nThere are a couple of recent papers published by Los Alamos National Laboratory [paper 1][4] and [paper 2][5]\n\nHope this is useful !!\n\n\n [1]: http://docs.obspy.org/\n [2]: http://docs.obspy.org/tutorial/index.html\n [3]: https://pyrocko.org/docs/current/\n [4]: https://www.nature.com/articles/s41561-018-0272-8\n [5]: https://www.nature.com/articles/s41561-018-0274-6", "votes": null }, { "id": "454541", "postDate": "01/11/2019 20:04:04", "content": "Thanks for the data. Is there any way to get free issues of the papers1/2 ?
", "rawMarkdown": "Thanks for the data. Is there any way to get free issues of the papers1/2 ?", "votes": null }, { "id": "454714", "postDate": "01/12/2019 04:33:29", "content": "Thank you for the tips, Vishy!
", "rawMarkdown": "Thank you for the tips, Vishy!", "votes": null }, { "id": "454761", "postDate": "01/12/2019 06:58:53", "content": "I can give you the papers if you contact me.
", "rawMarkdown": "I can give you the papers if you contact me.", "votes": null }, { "id": "454938", "postDate": "01/12/2019 15:01:41", "content": "Thank you @vishy
", "rawMarkdown": "Thank you @vishy", "votes": null }, { "id": "454970", "postDate": "01/12/2019 16:49:48", "content": "Thank you Vishy! Very helpful.
", "rawMarkdown": "Thank you Vishy! Very helpful.", "votes": null }, { "id": "455024", "postDate": "01/12/2019 19:19:39", "content": "Thanks for link. Very helpful. I'm trying to use some ideas from the\" ObsPy \" package. Let's see what happens
", "rawMarkdown": "Thanks for link. Very helpful. I'm trying to use some ideas from the\" ObsPy \" package. Let's see what happens", "votes": null }, { "id": "455113", "postDate": "01/13/2019 02:19:53", "content": "I am trying to understand the train.csv file. What does acoustic_data represent and how does it relate to time_to_failure? I could not find the it explained in data description.
\n\nThanks
", "rawMarkdown": "I am trying to understand the train.csv file. What does acoustic_data represent and how does it relate to time_to_failure? I could not find the it explained in data description.\n\nThanks", "votes": null }, { "id": "455129", "postDate": "01/13/2019 03:10:44", "content": "Thanks!
", "rawMarkdown": "Thanks!", "votes": null }, { "id": "455166", "postDate": "01/13/2019 05:35:35", "content": "Another way is also to register in ResearchGate and request author to share the Full-text. I have done from my end and would have to wait and see if they provide access. Will keep you all posted..
", "rawMarkdown": "Another way is also to register in ResearchGate and request author to share the Full-text. I have done from my end and would have to wait and see if they provide access. Will keep you all posted..", "votes": null }, { "id": "455167", "postDate": "01/13/2019 05:37:10", "content": "Sure please share as kernels is feasible..
", "rawMarkdown": "Sure please share as kernels is feasible..", "votes": null }, { "id": "455346", "postDate": "01/13/2019 16:28:43", "content": "Example use function for signal data https://www.kaggle.com/nikitagribov/analysis-function-for-seismic-signal-data
\n\nExample my public kernel https://www.kaggle.com/nikitagribov/seismic-signal-eda-analysis-function
", "rawMarkdown": "Example use function for signal data https://www.kaggle.com/nikitagribov/analysis-function-for-seismic-signal-data\n\nExample my public kernel https://www.kaggle.com/nikitagribov/seismic-signal-eda-analysis-function", "votes": null }, { "id": "455576", "postDate": "01/14/2019 07:19:33", "content": "Thanks much for sharing this, there are several learning from both of your kernels mentioned.. I guess newbies like me in Seismology would have a lot of learning by the end of the competition..
", "rawMarkdown": "Thanks much for sharing this, there are several learning from both of your kernels mentioned.. I guess newbies like me in Seismology would have a lot of learning by the end of the competition..", "votes": null }, { "id": "456318", "postDate": "01/15/2019 14:52:04", "content": "Thanks!
", "rawMarkdown": "Thanks!", "votes": null }, { "id": "456348", "postDate": "01/15/2019 16:30:14", "content": "Hello Ayan, would you please share these papers with me also! I am unable to message you directly
", "rawMarkdown": "Hello Ayan, would you please share these papers with me also! I am unable to message you directly", "votes": null }, { "id": "456543", "postDate": "01/16/2019 02:59:28", "content": "I'm confused about time to failure and the structure of the data in general. Each training file has 150000 records so I assume that each 150000 record chunk represents an observation in the training data. For each training observation there are 150000 different values but we are supposed to predict a single value for a 150000 record observation in the test set. Are we predicting the average? I'm probably missing something simple since nobody else is asking this.
", "rawMarkdown": "I'm confused about time to failure and the structure of the data in general. Each training file has 150000 records so I assume that each 150000 record chunk represents an observation in the training data. For each training observation there are 150000 different values but we are supposed to predict a single value for a 150000 record observation in the test set. Are we predicting the average? I'm probably missing something simple since nobody else is asking this.", "votes": null }, { "id": "456562", "postDate": "01/16/2019 03:57:03", "content": "The failure time, y, has been pretty much linearly interpolated from the beginning to the end of each earthquake.
\n\nA 150,000 sequence only represents 0.0375 sec span. \nConsidering an error of 0.0375, it is really nothing, for the best LB scores are still revolving around 1.5 sec error.
\n\nThat's also why the starter kernel uses the last value in the sequence as the single one outcome value.
", "rawMarkdown": "The failure time, y, has been pretty much linearly interpolated from the beginning to the end of each earthquake.\n\nA 150,000 sequence only represents 0.0375 sec span. \nConsidering an error of 0.0375, it is really nothing, for the best LB scores are still revolving around 1.5 sec error.\n\nThat's also why the starter kernel uses the last value in the sequence as the single one outcome value.", "votes": null }, { "id": "457930", "postDate": "01/18/2019 11:06:39", "content": "Unable to install obspy.. Could you please help me.
", "rawMarkdown": "Unable to install obspy.. Could you please help me.", "votes": null }, { "id": "458161", "postDate": "01/19/2019 00:06:09", "content": "There are instructions available at link. Could you also share a specific error message and the environment being used.
", "rawMarkdown": "There are instructions available at [link][1]. Could you also share a specific error message and the environment being used.\n\n\n [1]: https://github.com/obspy/obspy/wiki#installation", "votes": null }, { "id": "468050", "postDate": "02/08/2019 07:40:12", "content": "Thanks!
", "rawMarkdown": "Thanks!", "votes": null } ], "comments": [ { "id": 454541, "author_name": "georgem", "author_url": "", "post_date": "01/11/2019 20:04:04", "content": "Thanks for the data. Is there any way to get free issues of the papers1/2 ?
", "votes": null, "replies": [ { "id": 454761, "author_name": "ayansengupta17", "author_url": "", "post_date": "01/12/2019 06:58:53", "content": "I can give you the papers if you contact me.
", "votes": null, "replies": [] }, { "id": 455166, "author_name": "viswanathravindran", "author_url": "", "post_date": "01/13/2019 05:35:35", "content": "Another way is also to register in ResearchGate and request author to share the Full-text. I have done from my end and would have to wait and see if they provide access. Will keep you all posted..
", "votes": null, "replies": [] }, { "id": 456348, "author_name": "quinlivanb", "author_url": "", "post_date": "01/15/2019 16:30:14", "content": "Hello Ayan, would you please share these papers with me also! I am unable to message you directly
", "votes": null, "replies": [] } ] }, { "id": 454714, "author_name": "guilhermecaeiro", "author_url": "", "post_date": "01/12/2019 04:33:29", "content": "Thank you for the tips, Vishy!
", "votes": null, "replies": [] }, { "id": 454938, "author_name": "ramisharanik", "author_url": "", "post_date": "01/12/2019 15:01:41", "content": "Thank you @vishy
", "votes": null, "replies": [] }, { "id": 454970, "author_name": "jeremymaurer", "author_url": "", "post_date": "01/12/2019 16:49:48", "content": "Thank you Vishy! Very helpful.
", "votes": null, "replies": [] }, { "id": 455024, "author_name": "nikitagribov", "author_url": "", "post_date": "01/12/2019 19:19:39", "content": "Thanks for link. Very helpful. I'm trying to use some ideas from the\" ObsPy \" package. Let's see what happens
", "votes": null, "replies": [ { "id": 455167, "author_name": "viswanathravindran", "author_url": "", "post_date": "01/13/2019 05:37:10", "content": "Sure please share as kernels is feasible..
", "votes": null, "replies": [] }, { "id": 455346, "author_name": "nikitagribov", "author_url": "", "post_date": "01/13/2019 16:28:43", "content": "Example use function for signal data https://www.kaggle.com/nikitagribov/analysis-function-for-seismic-signal-data
\n\nExample my public kernel https://www.kaggle.com/nikitagribov/seismic-signal-eda-analysis-function
", "votes": null, "replies": [] }, { "id": 455576, "author_name": "viswanathravindran", "author_url": "", "post_date": "01/14/2019 07:19:33", "content": "Thanks much for sharing this, there are several learning from both of your kernels mentioned.. I guess newbies like me in Seismology would have a lot of learning by the end of the competition..
", "votes": null, "replies": [] } ] }, { "id": 455113, "author_name": "rmasoor", "author_url": "", "post_date": "01/13/2019 02:19:53", "content": "I am trying to understand the train.csv file. What does acoustic_data represent and how does it relate to time_to_failure? I could not find the it explained in data description.
\n\nThanks
", "votes": null, "replies": [] }, { "id": 455129, "author_name": "kramea", "author_url": "", "post_date": "01/13/2019 03:10:44", "content": "Thanks!
", "votes": null, "replies": [] }, { "id": 456318, "author_name": "nikolaym", "author_url": "", "post_date": "01/15/2019 14:52:04", "content": "Thanks!
", "votes": null, "replies": [] }, { "id": 456543, "author_name": "alkhwarizmi", "author_url": "", "post_date": "01/16/2019 02:59:28", "content": "I'm confused about time to failure and the structure of the data in general. Each training file has 150000 records so I assume that each 150000 record chunk represents an observation in the training data. For each training observation there are 150000 different values but we are supposed to predict a single value for a 150000 record observation in the test set. Are we predicting the average? I'm probably missing something simple since nobody else is asking this.
", "votes": null, "replies": [ { "id": 456562, "author_name": "tclf90", "author_url": "", "post_date": "01/16/2019 03:57:03", "content": "The failure time, y, has been pretty much linearly interpolated from the beginning to the end of each earthquake.
\n\nA 150,000 sequence only represents 0.0375 sec span. \nConsidering an error of 0.0375, it is really nothing, for the best LB scores are still revolving around 1.5 sec error.
\n\nThat's also why the starter kernel uses the last value in the sequence as the single one outcome value.
", "votes": null, "replies": [] } ] }, { "id": 457930, "author_name": "sasmitamohanty", "author_url": "", "post_date": "01/18/2019 11:06:39", "content": "Unable to install obspy.. Could you please help me.
", "votes": null, "replies": [ { "id": 458161, "author_name": "viswanathravindran", "author_url": "", "post_date": "01/19/2019 00:06:09", "content": "There are instructions available at link. Could you also share a specific error message and the environment being used.
", "votes": null, "replies": [] } ] }, { "id": 468050, "author_name": "stavya", "author_url": "", "post_date": "02/08/2019 07:40:12", "content": "Thanks!
", "votes": null, "replies": [] } ], "raw_markdown_by_id": { "453999": "Some quick tips for getting started with this competition:\n\nFound the \"ObsPy\" package which can handle seismology data - [link][1]\nSpecifically the ObsPy tutorial can be helpful to derive feature - [link][2] \nAnother Python package \"pyrocko\" for handling and understanding seismology data - [link][3]\nThere are a couple of recent papers published by Los Alamos National Laboratory [paper 1][4] and [paper 2][5]\n\nHope this is useful !!\n\n\n [1]: http://docs.obspy.org/\n [2]: http://docs.obspy.org/tutorial/index.html\n [3]: https://pyrocko.org/docs/current/\n [4]: https://www.nature.com/articles/s41561-018-0272-8\n [5]: https://www.nature.com/articles/s41561-018-0274-6", "454541": "Thanks for the data. Is there any way to get free issues of the papers1/2 ?", "454714": "Thank you for the tips, Vishy!", "454761": "I can give you the papers if you contact me.", "454938": "Thank you @vishy", "454970": "Thank you Vishy! Very helpful.", "455024": "Thanks for link. Very helpful. I'm trying to use some ideas from the\" ObsPy \" package. Let's see what happens", "455113": "I am trying to understand the train.csv file. What does acoustic_data represent and how does it relate to time_to_failure? I could not find the it explained in data description.\n\nThanks", "455129": "Thanks!", "455166": "Another way is also to register in ResearchGate and request author to share the Full-text. I have done from my end and would have to wait and see if they provide access. Will keep you all posted..", "455167": "Sure please share as kernels is feasible..", "455346": "Example use function for signal data https://www.kaggle.com/nikitagribov/analysis-function-for-seismic-signal-data\n\nExample my public kernel https://www.kaggle.com/nikitagribov/seismic-signal-eda-analysis-function", "455576": "Thanks much for sharing this, there are several learning from both of your kernels mentioned.. I guess newbies like me in Seismology would have a lot of learning by the end of the competition..", "456318": "Thanks!", "456348": "Hello Ayan, would you please share these papers with me also! I am unable to message you directly", "456543": "I'm confused about time to failure and the structure of the data in general. Each training file has 150000 records so I assume that each 150000 record chunk represents an observation in the training data. For each training observation there are 150000 different values but we are supposed to predict a single value for a 150000 record observation in the test set. Are we predicting the average? I'm probably missing something simple since nobody else is asking this.", "456562": "The failure time, y, has been pretty much linearly interpolated from the beginning to the end of each earthquake.\n\nA 150,000 sequence only represents 0.0375 sec span. \nConsidering an error of 0.0375, it is really nothing, for the best LB scores are still revolving around 1.5 sec error.\n\nThat's also why the starter kernel uses the last value in the sequence as the single one outcome value.", "457930": "Unable to install obspy.. Could you please help me.", "458161": "There are instructions available at [link][1]. Could you also share a specific error message and the environment being used.\n\n\n [1]: https://github.com/obspy/obspy/wiki#installation", "468050": "Thanks!" }, "source": "meta" }