Created
September 27, 2023 20:58
-
-
Save jbusecke/2cd85c532682cf3dd7e0b693bdb66cad to your computer and use it in GitHub Desktop.
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
{ | |
"cells": [ | |
{ | |
"cell_type": "code", | |
"execution_count": 1, | |
"id": "bb89df8d-3a0b-478c-a7fc-f846b673d819", | |
"metadata": { | |
"tags": [] | |
}, | |
"outputs": [], | |
"source": [ | |
"# !pip install pip install git+https://github.com/jbusecke/pangeo-forge-esgf.git@beam-refactor" | |
] | |
}, | |
{ | |
"cell_type": "code", | |
"execution_count": 2, | |
"id": "929af61e-0d0b-4a18-a1c7-7953d1c35faf", | |
"metadata": { | |
"tags": [] | |
}, | |
"outputs": [], | |
"source": [ | |
"import gcsfs\n", | |
"import zarr\n", | |
"from typing import List\n", | |
"cache_bucket = \"leap-scratch/data-library/cache\"\n", | |
"target_bucket = \"leap-persistent-ro/data-library/cmip6-testing\"\n", | |
"\n", | |
"\n", | |
"def get_job_details(jobid: str):\n", | |
" pass\n", | |
"\n", | |
"def get_cached_files(iid: str) -> List[str]:\n", | |
" fs = gcsfs.GCSFileSystem()\n", | |
" glob_wildcard = '*'.join(['']+iid.lower().split('.')[2:]+[''])\n", | |
" # this takes quite long...\n", | |
" all_files = fs.glob(f\"{cache_bucket}/{glob_wildcard}\")\n", | |
" return [f\"gs://{f}\" for f in all_files]\n", | |
"\n", | |
"def get_written_store(jobid: str, iid: str) -> str:\n", | |
" subfolder = '-'.join(jobid.split('-')[:-1])\n", | |
" return f\"gs://{target_bucket}/{subfolder}/{iid}.zarr\"" | |
] | |
}, | |
{ | |
"cell_type": "code", | |
"execution_count": 3, | |
"id": "48b89aa4-c348-4904-a67f-04a396ba0e18", | |
"metadata": { | |
"tags": [] | |
}, | |
"outputs": [], | |
"source": [ | |
"iids = ['CMIP6.ScenarioMIP.CSIRO-ARCCSS.ACCESS-CM2.ssp585.r1i1p1f1.day.pr.gn.v20210317',\n", | |
"'CMIP6.ScenarioMIP.CSIRO.ACCESS-ESM1-5.ssp585.r4i1p1f1.day.sfcWind.gn.v20210318',\n", | |
"'CMIP6.ScenarioMIP.CSIRO.ACCESS-ESM1-5.ssp585.r6i1p1f1.day.sfcWind.gn.v20210318',\n", | |
"'CMIP6.ScenarioMIP.CSIRO.ACCESS-ESM1-5.ssp585.r4i1p1f1.day.pr.gn.v20210318',\n", | |
"'CMIP6.ScenarioMIP.CSIRO.ACCESS-ESM1-5.ssp585.r10i1p1f1.day.pr.gn.v20210318',\n", | |
"'CMIP6.CMIP.MRI.MRI-ESM2-0.historical.r2i1p1f1.day.pr.gn.v20190603',\n", | |
"'CMIP6.CMIP.MRI.MRI-ESM2-0.historical.r3i1p1f1.day.pr.gn.v20190603',\n", | |
"'CMIP6.CMIP.MRI.MRI-ESM2-0.historical.r4i1p1f1.day.pr.gn.v20190603',\n", | |
"'CMIP6.CMIP.MRI.MRI-ESM2-0.historical.r5i1p1f1.day.pr.gn.v20190603',\n", | |
"]" | |
] | |
}, | |
{ | |
"cell_type": "code", | |
"execution_count": 4, | |
"id": "e892d48f-a0b1-42b1-9e8f-268fd4fe8bfe", | |
"metadata": { | |
"tags": [] | |
}, | |
"outputs": [ | |
{ | |
"name": "stdout", | |
"output_type": "stream", | |
"text": [ | |
"pangeo_forge_esgf.recipe_inputs - INFO - Checking responsiveness of search_nodes=['http://esgf-node.llnl.gov/esg-search/search', 'http://esgf-data.dkrz.de/esg-search/search', 'http://esgf-node.ipsl.upmc.fr/esg-search/search', 'http://esgf-index1.ceda.ac.uk/esg-search/search', 'http://esg-dn1.nsc.liu.se/esg-search/search', 'http://esgf.nci.org.au/esg-search/search']\n", | |
"pangeo_forge_esgf.recipe_inputs - INFO - responsive_search_nodes=['http://esgf-node.llnl.gov/esg-search/search', 'http://esgf-data.dkrz.de/esg-search/search', 'http://esgf-node.ipsl.upmc.fr/esg-search/search', 'http://esg-dn1.nsc.liu.se/esg-search/search', 'http://esgf.nci.org.au/esg-search/search']\n", | |
"pangeo_forge_esgf.recipe_inputs - INFO - Requesting urls\n" | |
] | |
}, | |
{ | |
"name": "stderr", | |
"output_type": "stream", | |
"text": [ | |
"100%|██████████| 45/45 [00:10<00:00, 4.25it/s]" | |
] | |
}, | |
{ | |
"name": "stdout", | |
"output_type": "stream", | |
"text": [ | |
"pangeo_forge_esgf.recipe_inputs - INFO - Processing responses\n", | |
"pangeo_forge_esgf.recipe_inputs - INFO - Processing responses: Expected files per iid\n", | |
"pangeo_forge_esgf.recipe_inputs - INFO - Processing responses: Check for missing iids\n", | |
"pangeo_forge_esgf.recipe_inputs - WARNING - Not able to find results for the following 4 iids: ['CMIP6.CMIP.MRI.MRI-ESM2-0.historical.r5i1p1f1.day.pr.gn.v20190603', 'CMIP6.CMIP.MRI.MRI-ESM2-0.historical.r3i1p1f1.day.pr.gn.v20190603', 'CMIP6.CMIP.MRI.MRI-ESM2-0.historical.r2i1p1f1.day.pr.gn.v20190603', 'CMIP6.CMIP.MRI.MRI-ESM2-0.historical.r4i1p1f1.day.pr.gn.v20190603']\n", | |
"pangeo_forge_esgf.recipe_inputs - INFO - Processing responses: Flatten results\n", | |
"pangeo_forge_esgf.recipe_inputs - INFO - Processing responses: Group results\n", | |
"pangeo_forge_esgf.recipe_inputs - INFO - Choosing one url per file\n", | |
"pangeo_forge_esgf.recipe_inputs - WARNING - This method seems to be unreliable for getting many urls. \n", | |
"If you are getting less datasets than you expect, try 'first' instead.\n", | |
"pangeo_forge_esgf.recipe_inputs - INFO - Find first responsive url for each file\n" | |
] | |
}, | |
{ | |
"name": "stderr", | |
"output_type": "stream", | |
"text": [ | |
"\n", | |
"100%|██████████| 8/8 [00:00<00:00, 12.98it/s]" | |
] | |
}, | |
{ | |
"name": "stdout", | |
"output_type": "stream", | |
"text": [ | |
"pangeo_forge_esgf.recipe_inputs - WARNING - Was not able to construct url list for (4/9) iids\n", | |
"pangeo_forge_esgf.recipe_inputs - INFO - Was not able to construct url list for the following iids:\n", | |
"pangeo_forge_esgf.recipe_inputs - INFO - {'CMIP6.CMIP.MRI.MRI-ESM2-0.historical.r5i1p1f1.day.pr.gn.v20190603', 'CMIP6.CMIP.MRI.MRI-ESM2-0.historical.r3i1p1f1.day.pr.gn.v20190603', 'CMIP6.CMIP.MRI.MRI-ESM2-0.historical.r2i1p1f1.day.pr.gn.v20190603', 'CMIP6.CMIP.MRI.MRI-ESM2-0.historical.r4i1p1f1.day.pr.gn.v20190603'}\n" | |
] | |
}, | |
{ | |
"name": "stderr", | |
"output_type": "stream", | |
"text": [ | |
"\n" | |
] | |
} | |
], | |
"source": [ | |
"# query pangeo-forge-esgf for the urls\n", | |
"# TODO: I need to write the order of urls into the attrs\n", | |
"import logging\n", | |
"import pangeo_forge_esgf\n", | |
"from pangeo_forge_esgf import get_urls_from_esgf, setup_logging\n", | |
"\n", | |
"# setup_logging('DEBUG')\n", | |
"setup_logging('INFO')\n", | |
"\n", | |
"url_dict = await get_urls_from_esgf(iids, limit_per_host=50, max_concurrency=50, choose_url='first_responsive')" | |
] | |
}, | |
{ | |
"cell_type": "code", | |
"execution_count": 12, | |
"id": "8f290281-ba3a-45cf-b436-112170ec3439", | |
"metadata": {}, | |
"outputs": [ | |
{ | |
"data": { | |
"text/plain": [ | |
"['CMIP6.ScenarioMIP.CSIRO-ARCCSS.ACCESS-CM2.ssp585.r1i1p1f1.day.pr.gn.v20210317',\n", | |
" 'CMIP6.ScenarioMIP.CSIRO.ACCESS-ESM1-5.ssp585.r4i1p1f1.day.sfcWind.gn.v20210318',\n", | |
" 'CMIP6.ScenarioMIP.CSIRO.ACCESS-ESM1-5.ssp585.r6i1p1f1.day.sfcWind.gn.v20210318',\n", | |
" 'CMIP6.ScenarioMIP.CSIRO.ACCESS-ESM1-5.ssp585.r4i1p1f1.day.pr.gn.v20210318',\n", | |
" 'CMIP6.ScenarioMIP.CSIRO.ACCESS-ESM1-5.ssp585.r10i1p1f1.day.pr.gn.v20210318']" | |
] | |
}, | |
"execution_count": 12, | |
"metadata": {}, | |
"output_type": "execute_result" | |
} | |
], | |
"source": [ | |
"iids_with_urls = list(url_dict.keys())\n", | |
"iids_with_urls" | |
] | |
}, | |
{ | |
"cell_type": "code", | |
"execution_count": 25, | |
"id": "b23750b9-1266-4e9d-ae49-e6a10ab75146", | |
"metadata": { | |
"tags": [] | |
}, | |
"outputs": [], | |
"source": [ | |
"# from ../cmip6-leap-feedstock/catalog_utils import convert_cmip6_df_to_iid_df\n", | |
"# FIXME: Import this properly\n", | |
"import pandas as pd\n", | |
"def _maybe_join(iterable):\n", | |
" assert len(iterable) == 2 \n", | |
" dcpp_init_year = iterable[0]\n", | |
" member_id = iterable[1]\n", | |
" if not pd.isnull(dcpp_init_year):\n", | |
" return f\"{dcpp_init_year}-{member_id}\"\n", | |
" else:\n", | |
" return member_id\n", | |
"\n", | |
"def convert_cmip6_df_to_iid_df(df: pd.DataFrame) -> pd.DataFrame:\n", | |
" # now remove the ones already in the pangeo catalog\n", | |
"\n", | |
" df['variant_label'] = df[['dcpp_init_year', 'member_id']].agg(_maybe_join, axis=1)\n", | |
" df['version'] = 'v'+df['version'].astype(str)\n", | |
" df['instance_id'] = df[['activity_id', 'institution_id', 'source_id', 'experiment_id',\n", | |
" 'variant_label', 'table_id', 'variable_id', 'grid_label', 'version']].astype(str).agg('.'.join, axis=1).tolist()\n", | |
" df['instance_id'] = 'CMIP6.'+df['instance_id']\n", | |
" df['store'] = df['zstore']\n", | |
" # add current time as bigquery timestamp\n", | |
" df['timestamp'] = pd.Timestamp.now(tz='UTC')\n", | |
" df = df[['instance_id','store', 'timestamp']]\n", | |
" return df" | |
] | |
}, | |
{ | |
"cell_type": "code", | |
"execution_count": 34, | |
"id": "6c7f4e29-cbed-4ebd-9e49-9e2e3b76d8e1", | |
"metadata": {}, | |
"outputs": [ | |
{ | |
"data": { | |
"text/plain": [ | |
"['CMIP6.HighResMIP.CMCC.CMCC-CM2-HR4.highresSST-present.r1i1p1f1.Amon.ps.gn.v20170706',\n", | |
" 'CMIP6.HighResMIP.CMCC.CMCC-CM2-HR4.highresSST-present.r1i1p1f1.Amon.rsds.gn.v20170706',\n", | |
" 'CMIP6.HighResMIP.CMCC.CMCC-CM2-HR4.highresSST-present.r1i1p1f1.Amon.rlus.gn.v20170706',\n", | |
" 'CMIP6.HighResMIP.CMCC.CMCC-CM2-HR4.highresSST-present.r1i1p1f1.Amon.rlds.gn.v20170706',\n", | |
" 'CMIP6.HighResMIP.CMCC.CMCC-CM2-HR4.highresSST-present.r1i1p1f1.Amon.psl.gn.v20170706',\n", | |
" 'CMIP6.HighResMIP.CMCC.CMCC-CM2-HR4.highresSST-present.r1i1p1f1.Amon.hurs.gn.v20170706',\n", | |
" 'CMIP6.HighResMIP.CMCC.CMCC-CM2-HR4.highresSST-present.r1i1p1f1.Amon.huss.gn.v20170706',\n", | |
" 'CMIP6.HighResMIP.CMCC.CMCC-CM2-HR4.highresSST-present.r1i1p1f1.Amon.hus.gn.v20170706',\n", | |
" 'CMIP6.HighResMIP.CMCC.CMCC-CM2-HR4.highresSST-present.r1i1p1f1.Amon.hfss.gn.v20170706',\n", | |
" 'CMIP6.HighResMIP.CMCC.CMCC-CM2-HR4.highresSST-present.r1i1p1f1.Amon.rsus.gn.v20170706']" | |
] | |
}, | |
"execution_count": 34, | |
"metadata": {}, | |
"output_type": "execute_result" | |
} | |
], | |
"source": [ | |
"import intake\n", | |
"# col = intake.open_esm_datastore(\n", | |
"# \"https://storage.googleapis.com/leap-persistent-ro/data-library/catalogs/cmip6-test/leap-pangeo-cmip6-noqc-test.json\"\n", | |
"# )\n", | |
"from xmip.utils import google_cmip_col\n", | |
"col = google_cmip_col()\n", | |
"\n", | |
"df = convert_cmip6_df_to_iid_df(col.df)\n", | |
"iids_in_catalog = df['instance_id'].tolist()\n", | |
"iids_in_catalog[:10]" | |
] | |
}, | |
{ | |
"cell_type": "code", | |
"execution_count": 36, | |
"id": "e25a3deb-7f75-4708-b03c-732a05f2a76a", | |
"metadata": { | |
"tags": [] | |
}, | |
"outputs": [ | |
{ | |
"data": { | |
"text/plain": [ | |
"set()" | |
] | |
}, | |
"execution_count": 36, | |
"metadata": {}, | |
"output_type": "execute_result" | |
} | |
], | |
"source": [ | |
"set(iids_with_urls)-set(iids_in_catalog)" | |
] | |
}, | |
{ | |
"cell_type": "markdown", | |
"id": "f1dddfb3-fa97-4407-8dc2-4f62341c3b6e", | |
"metadata": { | |
"tags": [] | |
}, | |
"source": [ | |
"So these are all in the old catalog!" | |
] | |
}, | |
{ | |
"cell_type": "code", | |
"execution_count": null, | |
"id": "51ee2a2f-a3a3-4a36-baf4-4fd0ff94e632", | |
"metadata": {}, | |
"outputs": [], | |
"source": [] | |
} | |
], | |
"metadata": { | |
"kernelspec": { | |
"display_name": "Python 3 (ipykernel)", | |
"language": "python", | |
"name": "python3" | |
}, | |
"language_info": { | |
"codemirror_mode": { | |
"name": "ipython", | |
"version": 3 | |
}, | |
"file_extension": ".py", | |
"mimetype": "text/x-python", | |
"name": "python", | |
"nbconvert_exporter": "python", | |
"pygments_lexer": "ipython3", | |
"version": "3.10.12" | |
} | |
}, | |
"nbformat": 4, | |
"nbformat_minor": 5 | |
} |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment