%matplotlib inline
import pandas as pd
import socket
host = socket.getfqdn()
from core import load, zoom, calc, save,plots,monitor
#reload funcs after updating ./core/*.py
import importlib
importlib.reload(load)
importlib.reload(zoom)
importlib.reload(calc)
importlib.reload(save)
importlib.reload(plots)
importlib.reload(monitor)
<module 'core.monitor' from '/ccc/work/cont003/gen7420/odakatin/monitor-sedna/notebook/core/monitor.py'>
If you submit the job with job scheduler; below are list of enviroment variable one can pass
local : if True run dask local cluster, if not true, put number of workers setted in the 'local' if no 'local ' given, local will be setted automatically to 'True'
%env ychunk='2', #%env tchunk='2'
controls chunk. 'False' sets no modification from original netcdf file's chunk.
ychunk=10 will group the original netcdf file to 10 by 10
tchunk=1 will chunk the time coordinate one by one
%env file_exp=
'file_exp': Which 'experiment' name is it? this corresopnds to intake catalog name without path and .yaml
#%env year=
for Validation, this correspoinds to path/year/month 's year for monitoring, this corresponids to 'date' having means do all files in the monitoring directory setting it as 0[0-9] &1[0-9]& [2-3][0-9], the job can be separated in three lots. For DELTA experiment, year corresponds to really 'year'
%env month=
for monitoring this corresponds to file path path-XIOS.{month}/
For DELTA experiment, year corresponds to really 'month'
proceed saving? True or False , Default is setted as True
proceed plotting? True or False , Default is setted as True
proceed computation? or just load computed result? True or False , Default is setted as True
save output file used for plotting
using kerchunked file -> False, not using kerhcunk -> True
name of control file to be used for computation/plots/save/ We have number of M_xxx.csv
Monitor.sh calls M_MLD_2D
and AWTD.sh, Fluxnet.sh, Siconc.sh, IceClim.sh, FWC_SSH.sh, Integrals.sh , Sections.sh
M_AWTMD
M_Fluxnet
M_Ice_quantities
M_IceClim M_IceConce M_IceThick
M_FWC_2D M_FWC_integrals M_FWC_SSH M_SSH_anomaly
M_Mean_temp_velo M_Mooring
M_Sectionx M_Sectiony
%%time
# 'savefig': Do we save output in html? or not. keep it true.
savefig=True
client,cluster,control,catalog_url,month,year,daskreport,outputpath = load.set_control(host)
!mkdir -p $outputpath
!mkdir -p $daskreport
client
local True using host= irene5878.c-irene.mg1.tgcc.ccc.cea.fr starting dask cluster on local= True workers 16 10000000000 rome local cluster starting This code is running on irene5878.c-irene.mg1.tgcc.ccc.cea.fr using SEDNA_DELTA_MONITOR file experiment, read from ../lib/SEDNA_DELTA_MONITOR.yaml on year= 2012 on month= 03 outputpath= ../results/SEDNA_DELTA_MONITOR/ daskreport= ../results/dask/6419623irene5878.c-irene.mg1.tgcc.ccc.cea.fr_SEDNA_DELTA_MONITOR_03M_Sectiony/ CPU times: user 591 ms, sys: 130 ms, total: 721 ms Wall time: 23.8 s
Client-9e99a67e-13e4-11ed-9a41-080038b933dd
Connection method: Cluster object | Cluster type: distributed.LocalCluster |
Dashboard: http://127.0.0.1:8787/status |
be630af8
Dashboard: http://127.0.0.1:8787/status | Workers: 16 |
Total threads: 128 | Total memory: 251.06 GiB |
Status: running | Using processes: True |
Scheduler-7a4c9620-8189-4643-b35d-ffd255e919a4
Comm: tcp://127.0.0.1:38222 | Workers: 16 |
Dashboard: http://127.0.0.1:8787/status | Total threads: 128 |
Started: Just now | Total memory: 251.06 GiB |
Comm: tcp://127.0.0.1:38246 | Total threads: 8 |
Dashboard: http://127.0.0.1:36000/status | Memory: 15.69 GiB |
Nanny: tcp://127.0.0.1:41788 | |
Local directory: /tmp/dask-worker-space/worker-w_rsncgp |
Comm: tcp://127.0.0.1:42412 | Total threads: 8 |
Dashboard: http://127.0.0.1:46803/status | Memory: 15.69 GiB |
Nanny: tcp://127.0.0.1:33944 | |
Local directory: /tmp/dask-worker-space/worker-4xvfuin2 |
Comm: tcp://127.0.0.1:35961 | Total threads: 8 |
Dashboard: http://127.0.0.1:41278/status | Memory: 15.69 GiB |
Nanny: tcp://127.0.0.1:37977 | |
Local directory: /tmp/dask-worker-space/worker-panfi_wz |
Comm: tcp://127.0.0.1:45156 | Total threads: 8 |
Dashboard: http://127.0.0.1:36425/status | Memory: 15.69 GiB |
Nanny: tcp://127.0.0.1:45775 | |
Local directory: /tmp/dask-worker-space/worker-bw626hxw |
Comm: tcp://127.0.0.1:42466 | Total threads: 8 |
Dashboard: http://127.0.0.1:39814/status | Memory: 15.69 GiB |
Nanny: tcp://127.0.0.1:33901 | |
Local directory: /tmp/dask-worker-space/worker-8995l2jt |
Comm: tcp://127.0.0.1:45682 | Total threads: 8 |
Dashboard: http://127.0.0.1:43132/status | Memory: 15.69 GiB |
Nanny: tcp://127.0.0.1:41862 | |
Local directory: /tmp/dask-worker-space/worker-z1_g40c2 |
Comm: tcp://127.0.0.1:33998 | Total threads: 8 |
Dashboard: http://127.0.0.1:40054/status | Memory: 15.69 GiB |
Nanny: tcp://127.0.0.1:33508 | |
Local directory: /tmp/dask-worker-space/worker-za3n9gkn |
Comm: tcp://127.0.0.1:42205 | Total threads: 8 |
Dashboard: http://127.0.0.1:36952/status | Memory: 15.69 GiB |
Nanny: tcp://127.0.0.1:34058 | |
Local directory: /tmp/dask-worker-space/worker-apv0cam4 |
Comm: tcp://127.0.0.1:35855 | Total threads: 8 |
Dashboard: http://127.0.0.1:32921/status | Memory: 15.69 GiB |
Nanny: tcp://127.0.0.1:40172 | |
Local directory: /tmp/dask-worker-space/worker-9xshskih |
Comm: tcp://127.0.0.1:37479 | Total threads: 8 |
Dashboard: http://127.0.0.1:39901/status | Memory: 15.69 GiB |
Nanny: tcp://127.0.0.1:38237 | |
Local directory: /tmp/dask-worker-space/worker-d2gxphr4 |
Comm: tcp://127.0.0.1:41529 | Total threads: 8 |
Dashboard: http://127.0.0.1:41210/status | Memory: 15.69 GiB |
Nanny: tcp://127.0.0.1:33302 | |
Local directory: /tmp/dask-worker-space/worker-nx79_r66 |
Comm: tcp://127.0.0.1:43324 | Total threads: 8 |
Dashboard: http://127.0.0.1:44797/status | Memory: 15.69 GiB |
Nanny: tcp://127.0.0.1:37206 | |
Local directory: /tmp/dask-worker-space/worker-eb8bkgbs |
Comm: tcp://127.0.0.1:41105 | Total threads: 8 |
Dashboard: http://127.0.0.1:42248/status | Memory: 15.69 GiB |
Nanny: tcp://127.0.0.1:40010 | |
Local directory: /tmp/dask-worker-space/worker-ovh5uckb |
Comm: tcp://127.0.0.1:33900 | Total threads: 8 |
Dashboard: http://127.0.0.1:37420/status | Memory: 15.69 GiB |
Nanny: tcp://127.0.0.1:35145 | |
Local directory: /tmp/dask-worker-space/worker-6g729aeh |
Comm: tcp://127.0.0.1:34908 | Total threads: 8 |
Dashboard: http://127.0.0.1:33365/status | Memory: 15.69 GiB |
Nanny: tcp://127.0.0.1:43077 | |
Local directory: /tmp/dask-worker-space/worker-yo2r3zul |
Comm: tcp://127.0.0.1:34125 | Total threads: 8 |
Dashboard: http://127.0.0.1:42493/status | Memory: 15.69 GiB |
Nanny: tcp://127.0.0.1:46708 | |
Local directory: /tmp/dask-worker-space/worker-iuk0gehz |
df=load.controlfile(control)
#Take out 'later' tagged computations
#df=df[~df['Value'].str.contains('later')]
df
Value | Inputs | Equation | Zone | Plot | Colourmap | MinMax | Unit | Oldname | Unnamed: 10 | |
---|---|---|---|---|---|---|---|---|---|---|
Section | gridT.votemper,gridS.vosaline,gridV.vomecrty,p... | data.unify_chunks().persist() | FramS | section | None | {'vosaline': (33,36.2), 'votemper': (-2,6), 'v... | None | S-1 |
Each computation consists of
%%time
import os
calcswitch=os.environ.get('calc', 'True')
lazy=os.environ.get('lazy','False' )
loaddata=((df.Inputs != '').any())
print('calcswitch=',calcswitch,'df.Inputs != nothing',loaddata, 'lazy=',lazy)
data = load.datas(catalog_url,df.Inputs,month,year,daskreport,lazy=lazy) if ((calcswitch=='True' )*loaddata) else 0
data
calcswitch= True df.Inputs != nothing True lazy= False ../lib/SEDNA_DELTA_MONITOR.yaml using param_xios reading ../lib/SEDNA_DELTA_MONITOR.yaml using param_xios reading <bound method DataSourceBase.describe of sources: param_xios: args: combine: nested concat_dim: y urlpath: /ccc/work/cont003/gen7420/odakatin/CONFIGS/SEDNA/SEDNA-I/SEDNA_Domain_cfg_Tgt_20210423_tsh10m_L1/param_f32/x_*.nc xarray_kwargs: compat: override coords: minimal data_vars: minimal parallel: true description: SEDNA NEMO parameters from MPI output nav_lon lat fails driver: intake_xarray.netcdf.NetCDFSource metadata: catalog_dir: /ccc/work/cont003/gen7420/odakatin/monitor-sedna/notebook/../lib/ > {'name': 'param_xios', 'container': 'xarray', 'plugin': ['netcdf'], 'driver': ['netcdf'], 'description': 'SEDNA NEMO parameters from MPI output nav_lon lat fails', 'direct_access': 'forbid', 'user_parameters': [{'name': 'path', 'description': 'file coordinate', 'type': 'str', 'default': '/ccc/work/cont003/gen7420/odakatin/CONFIGS/SEDNA/MESH/SEDNA_mesh_mask_Tgt_20210423_tsh10m_L1/param'}], 'metadata': {}, 'args': {'urlpath': '/ccc/work/cont003/gen7420/odakatin/CONFIGS/SEDNA/SEDNA-I/SEDNA_Domain_cfg_Tgt_20210423_tsh10m_L1/param_f32/x_*.nc', 'combine': 'nested', 'concat_dim': 'y'}} 0 read gridS ['vosaline'] lazy= False using load_data_xios_kerchunk reading gridS using load_data_xios_kerchunk reading <bound method DataSourceBase.describe of sources: data_xios_kerchunk: args: consolidated: false storage_options: fo: file:////ccc/cont003/home/ra5563/ra5563/catalogue/DELTA/201203/gridS_0[0-5][0-9][0-9].json target_protocol: file urlpath: reference:// description: CREG025 NEMO outputs from different xios server in kerchunk format driver: intake_xarray.xzarr.ZarrSource metadata: catalog_dir: /ccc/work/cont003/gen7420/odakatin/monitor-sedna/notebook/../lib/ > took 21.14095187187195 seconds 0 merging gridS ['vosaline'] 1 read gridT ['votemper'] lazy= False using load_data_xios_kerchunk reading gridT using load_data_xios_kerchunk reading <bound method DataSourceBase.describe of sources: data_xios_kerchunk: args: consolidated: false storage_options: fo: file:////ccc/cont003/home/ra5563/ra5563/catalogue/DELTA/201203/gridT_0[0-5][0-9][0-9].json target_protocol: file urlpath: reference:// description: CREG025 NEMO outputs from different xios server in kerchunk format driver: intake_xarray.xzarr.ZarrSource metadata: catalog_dir: /ccc/work/cont003/gen7420/odakatin/monitor-sedna/notebook/../lib/ > took 19.82506799697876 seconds 1 merging gridT ['votemper'] took 0.8125791549682617 seconds 2 read gridV ['vomecrty'] lazy= False using load_data_xios_kerchunk reading gridV using load_data_xios_kerchunk reading <bound method DataSourceBase.describe of sources: data_xios_kerchunk: args: consolidated: false storage_options: fo: file:////ccc/cont003/home/ra5563/ra5563/catalogue/DELTA/201203/gridV_0[0-5][0-9][0-9].json target_protocol: file urlpath: reference:// description: CREG025 NEMO outputs from different xios server in kerchunk format driver: intake_xarray.xzarr.ZarrSource metadata: catalog_dir: /ccc/work/cont003/gen7420/odakatin/monitor-sedna/notebook/../lib/ > took 23.905972719192505 seconds 2 merging gridV ['vomecrty'] took 0.8085231781005859 seconds param mask will be included in data param mask2d will be included in data param nav_lat will be included in data param nav_lon will be included in data param depth will be included in data CPU times: user 52.4 s, sys: 7.06 s, total: 59.5 s Wall time: 1min 31s
<xarray.Dataset> Dimensions: (t: 31, z: 150, y: 6540, x: 6560) Coordinates: time_centered (t) object dask.array<chunksize=(1,), meta=np.ndarray> * t (t) object 2012-03-01 12:00:00 ... 2012-03-31 12:00:00 * y (y) int64 1 2 3 4 5 6 7 ... 6535 6536 6537 6538 6539 6540 * x (x) int64 1 2 3 4 5 6 7 ... 6555 6556 6557 6558 6559 6560 * z (z) int64 1 2 3 4 5 6 7 8 ... 143 144 145 146 147 148 149 150 mask (z, y, x) bool dask.array<chunksize=(150, 13, 6560), meta=np.ndarray> mask2d (y, x) bool dask.array<chunksize=(13, 6560), meta=np.ndarray> nav_lat (y, x) float32 dask.array<chunksize=(13, 6560), meta=np.ndarray> nav_lon (y, x) float32 dask.array<chunksize=(13, 6560), meta=np.ndarray> depth (z, y, x) float32 dask.array<chunksize=(150, 13, 6560), meta=np.ndarray> Data variables: vosaline (t, z, y, x) float32 dask.array<chunksize=(1, 150, 13, 6560), meta=np.ndarray> votemper (t, z, y, x) float32 dask.array<chunksize=(1, 150, 13, 6560), meta=np.ndarray> vomecrty (t, z, y, x) float32 dask.array<chunksize=(1, 150, 13, 6560), meta=np.ndarray> Attributes: (12/26) CASE: DELTA CONFIG: SEDNA Conventions: CF-1.6 DOMAIN_dimensions_ids: [2, 3] DOMAIN_halo_size_end: [0, 0] DOMAIN_halo_size_start: [0, 0] ... ... nj: 13 output_frequency: 1d start_date: 20090101 timeStamp: 2022-Jan-19 12:40:34 GMT title: ocean T grid variables uuid: b2000bef-0e1f-4f0d-ae2a-b9429429b7a2
%%time
monitor.auto(df,data,savefig,daskreport,outputpath,file_exp='SEDNA'
)
#calc= True #save= True #plot= False Value='Section' Zone='FramS' Plot='section' cmap='None' clabel='None' clim= {'vosaline': (33, 36.2), 'votemper': (-2, 6), 'vomecrty': (-0.05, 0.05)} outputpath='../results/SEDNA_DELTA_MONITOR/' nc_outputpath='../nc_results/SEDNA_DELTA_MONITOR/' filename='SEDNA_section_FramS_Section' data=monitor.optimize_dataset(data) #2 Zooming Data data= zoom.FramS(data) data=monitor.optimize_dataset(data)
<xarray.Dataset> Dimensions: (t: 31, z: 150, x: 601) Coordinates: time_centered (t) object dask.array<chunksize=(1,), meta=np.ndarray> * t (t) object 2012-03-01 12:00:00 ... 2012-03-31 12:00:00 y int64 2609 * x (x) int64 3734 3735 3736 3737 3738 ... 4331 4332 4333 4334 * z (z) int64 1 2 3 4 5 6 7 8 ... 143 144 145 146 147 148 149 150 mask (z, x) bool dask.array<chunksize=(150, 601), meta=np.ndarray> mask2d (x) bool dask.array<chunksize=(601,), meta=np.ndarray> nav_lat (x) float32 dask.array<chunksize=(601,), meta=np.ndarray> nav_lon (x) float32 dask.array<chunksize=(601,), meta=np.ndarray> depth (z, x) float32 dask.array<chunksize=(150, 601), meta=np.ndarray> Data variables: vosaline (t, z, x) float32 dask.array<chunksize=(1, 150, 601), meta=np.ndarray> votemper (t, z, x) float32 dask.array<chunksize=(1, 150, 601), meta=np.ndarray> vomecrty (t, z, x) float32 dask.array<chunksize=(1, 150, 601), meta=np.ndarray> Attributes: (12/26) CASE: DELTA CONFIG: SEDNA Conventions: CF-1.6 DOMAIN_dimensions_ids: [2, 3] DOMAIN_halo_size_end: [0, 0] DOMAIN_halo_size_start: [0, 0] ... ... nj: 13 output_frequency: 1d start_date: 20090101 timeStamp: 2022-Jan-19 12:40:34 GMT title: ocean T grid variables uuid: b2000bef-0e1f-4f0d-ae2a-b9429429b7a2
#3 Start computing data= data.unify_chunks().persist() monitor.optimize_dataset(data) add optimise here once otimise can recognise
<xarray.Dataset> Dimensions: (t: 31, z: 150, x: 601) Coordinates: time_centered (t) object dask.array<chunksize=(1,), meta=np.ndarray> * t (t) object 2012-03-01 12:00:00 ... 2012-03-31 12:00:00 y int64 2609 * x (x) int64 3734 3735 3736 3737 3738 ... 4331 4332 4333 4334 * z (z) int64 1 2 3 4 5 6 7 8 ... 143 144 145 146 147 148 149 150 mask (z, x) bool dask.array<chunksize=(150, 601), meta=np.ndarray> mask2d (x) bool dask.array<chunksize=(601,), meta=np.ndarray> nav_lat (x) float32 dask.array<chunksize=(601,), meta=np.ndarray> nav_lon (x) float32 dask.array<chunksize=(601,), meta=np.ndarray> depth (z, x) float32 dask.array<chunksize=(150, 601), meta=np.ndarray> Data variables: vosaline (t, z, x) float32 dask.array<chunksize=(1, 150, 601), meta=np.ndarray> votemper (t, z, x) float32 dask.array<chunksize=(1, 150, 601), meta=np.ndarray> vomecrty (t, z, x) float32 dask.array<chunksize=(1, 150, 601), meta=np.ndarray> Attributes: (12/26) CASE: DELTA CONFIG: SEDNA Conventions: CF-1.6 DOMAIN_dimensions_ids: [2, 3] DOMAIN_halo_size_end: [0, 0] DOMAIN_halo_size_start: [0, 0] ... ... nj: 13 output_frequency: 1d start_date: 20090101 timeStamp: 2022-Jan-19 12:40:34 GMT title: ocean T grid variables uuid: b2000bef-0e1f-4f0d-ae2a-b9429429b7a2
#4 Saving SEDNA_section_FramS_Section data=save.datas(data,plot=Plot,path=nc_outputpath,filename=filename) start saving data saving data in a file t (1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1) 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 slice(0, 1, None) slice(1, 2, None) slice(2, 3, None) slice(3, 4, None) slice(4, 5, None) slice(5, 6, None) slice(6, 7, None) slice(7, 8, None) slice(8, 9, None) slice(9, 10, None) slice(10, 11, None) slice(11, 12, None) slice(12, 13, None) slice(13, 14, None) slice(14, 15, None) slice(15, 16, None) slice(16, 17, None) slice(17, 18, None) slice(18, 19, None) slice(19, 20, None) slice(20, 21, None) slice(21, 22, None) slice(22, 23, None) slice(23, 24, None) slice(24, 25, None) slice(25, 26, None) slice(26, 27, None) slice(27, 28, None) slice(28, 29, None) slice(29, 30, None) slice(30, 31, None) CPU times: user 7.91 s, sys: 779 ms, total: 8.69 s Wall time: 37.7 s