In the Tutorial 0: Noise generation i think there is a issue that create gaps in the data. Where there is:
\n\n\n\n
if segment > 0:
\nwriter_kwargs[\"tstart\"] += writer_kwargs[\"Tsft\"] + segment_lengths[segment - 1]
it must be:
\n\n\n\n
if segment > 0:
\nwriter_kwargs[\"tstart\"] += segment_lengths[segment - 1]
Otherwise, it creates gaps between data of Tsft length first, two times Tsft after, three times Tsft, and so on…
Supouse that writer_kwargs[\"tstart\"] starts at 0 and supouse that segment_lengths[0] and Tsftare equal 1. For the first fourier_data segment we'll have only one column and we want that the following segment starts at writer_kwargs[\"tstart\"] equal 1, not 2, as it works in the tutorial.
\nIf we have an other segment following the last one, assuming that the length is equal 1 for this, we want writer_kwargs[\"tstart\"] to be 2, not 4, as it working in the tutorial.
In the Tutorial 0: Noise generation i think there is a issue that create gaps in the data. Where there is:
\n\n\n\n
if segment > 0:
\nwriter_kwargs[\"tstart\"] += writer_kwargs[\"Tsft\"] + segment_lengths[segment - 1]
it must be:
\n\n\n\n
if segment > 0:
\nwriter_kwargs[\"tstart\"] += segment_lengths[segment - 1]
Otherwise, it creates gaps between data of Tsft length first, two times Tsft after, three times Tsft, and so on…
Supouse that writer_kwargs[\"tstart\"] starts at 0 and supouse that segment_lengths[0] and Tsftare equal 1. For the first fourier_data segment we'll have only one column and we want that the following segment starts at writer_kwargs[\"tstart\"] equal 1, not 2, as it works in the tutorial.
\nIf we have an other segment following the last one, assuming that the length is equal 1 for this, we want writer_kwargs[\"tstart\"] to be 2, not 4, as it working in the tutorial.
Hi @alexfernandezcz ,
\nIt's not necessarily incorrect, I just chose that convention back when I did the tutorials.
\nSince the section where this pops up is though to illustrate non-stationary noise rather than gaps
\nI don't think it's too much of an issue, but of course any suggestions are welcome as PRs.
Hi Rodrigo. In the tutorial it shows how to generate non-stationary noise, however it doesn't show how to incorporate signal into that non-stationary noise. Is there a tutorial somewhere I can find to do that?
", "rawMarkdown": "Hi Rodrigo. In the tutorial it shows how to generate non-stationary noise, however it doesn't show how to incorporate signal into that non-stationary noise. Is there a tutorial somewhere I can find to do that?", "votes": 1, "isDeleted": true }, { "id": 2038100, "postDate": "2022-11-21T07:40:27.230Z", "content": "Generate noise, save it as SFTs, provide the link to the saved file with noiseSFTs option of the writer.
\nhttps://pyfstat.readthedocs.io/en/latest/pyfstat.html#module-pyfstat.make_sfts
thank you))
", "rawMarkdown": "thank you))", "isDeleted": true } ] } ], "comments": [ { "id": 2034703, "author_name": "Rodrigo Tenorio", "author_url": "", "post_date": "2022-11-18T11:06:27.567000", "content": "Hi @alexfernandezcz ,
\nIt's not necessarily incorrect, I just chose that convention back when I did the tutorials.
\nSince the section where this pops up is though to illustrate non-stationary noise rather than gaps
\nI don't think it's too much of an issue, but of course any suggestions are welcome as PRs.
Hi Rodrigo. In the tutorial it shows how to generate non-stationary noise, however it doesn't show how to incorporate signal into that non-stationary noise. Is there a tutorial somewhere I can find to do that?
", "votes": 1, "replies": [] }, { "id": 2038100, "author_name": "DennisSakva", "author_url": "", "post_date": "2022-11-21T07:40:27.230000", "content": "Generate noise, save it as SFTs, provide the link to the saved file with noiseSFTs option of the writer.
\nhttps://pyfstat.readthedocs.io/en/latest/pyfstat.html#module-pyfstat.make_sfts
thank you))
", "votes": 0, "replies": [] } ] } ], "raw_markdown_by_id": { "2032664": "In the [Tutorial 0: Noise generation](https://github.com/PyFstat/PyFstat/blob/5ded4a672c3ab7309b64ce2531bcf704d4ed2f4e/examples/tutorials/0_generating_noise.ipynb) i think there is a issue that create gaps in the data. Where there is: \n\n>`if segment > 0:`\n>` writer_kwargs[\"tstart\"] += writer_kwargs[\"Tsft\"] + segment_lengths[segment - 1]`\n\nit must be:\n\n>`if segment > 0:`\n>` writer_kwargs[\"tstart\"] += segment_lengths[segment - 1]`\n\nOtherwise, it creates gaps between data of `Tsft` length first, two times `Tsft` after, three times `Tsft`, and so on... \n\nSupouse that `writer_kwargs[\"tstart\"]` starts at 0 and supouse that `segment_lengths[0]` and `Tsft`are equal 1. For the first `fourier_data` segment we'll have only one column and we want that the following segment starts at `writer_kwargs[\"tstart\"]` equal 1, not 2, as it works in the tutorial. \nIf we have an other segment following the last one, assuming that the length is equal 1 for this, we want `writer_kwargs[\"tstart\"]` to be 2, not 4, as it working in the tutorial. ", "2034703": "Hi @alexfernandezcz ,\n\nIt's not necessarily incorrect, I just chose that convention back when I did the tutorials.\nSince the section where this pops up is though to illustrate non-stationary noise rather than gaps\nI don't think it's too much of an issue, but of course any suggestions are welcome as PRs.\n\n" } }