Work plan PMP 2017/2018 #4

eguil · 2017-07-27T13:23:13Z

ENSO metrics notes 26/7/2017
Discussion:

design modular system for ENSO metrics that allows separation of concern and governance by appropriate experts
build on tests done at PCMDI on Bellenger et al. paper by Ji Woo and Peter

Ideas:

list of CLIVAR ENSO approved command lines to run enso_bellenger_compute.py (https://github.com/PCMDI/pcmdi_metrics/tree/issue_321_enso_bellenger_2017/src/python/devel/enso_bellenger)
develop driver_enso_metrics.py from enso_bellenger_compute.py API (Ji Woo to advise, Yann to test in ENSO metrics environment)
further develop enso_metrics_lib.py on github from programs by Ji Woo et al., further documenting the science - Yann and interface with ENSO CLIVAR
interface with CDAT/PMP libraries by Ji Woo and Peter
interface between CLIVAR ENSO science and metrics lib by Yann and Eric
provide reference json/graphics/table for past CMIPs : Ji Woo and Peter

Actions, timeline (and who):

telco late August to firm up work plan and timeline (Yann to organise)
produce first draft of work plan early Sept. (Eric to lead)
review by CLIVAR RF mid-Sept. (Andrew and Eric to lead)
implement first version by late Sept. (Yann and Ji Woo to lead)
present at Busan CLIVAR ENSO RF mid Oct (Yann and Eric)
review science contents and make final public release Nov 2017

eguil · 2017-08-31T17:06:38Z

Meeting Jiwoo/Yann/Eric on Aug 31st:

see slides 8 of
model_eval_governance_guilyardi_may17.pptx
Action Eric: access github x 3
Update enso_metrics_lib.py with more metrics from enso_bellenger_compute.py (Action: Yann and Eric)
Create a metrics collection (Action: Yann and Eric)
Modify enso_bellenger_compute.py into a driver_enso_metrics.py (Action Jiwoo)
Create a init_enso_metrics_obs.py (check if right data is there or download) (run by IT driver)
- Args: MCi, local_obs_path
- Check against a enso_ref_obs.py/json listing var/obs_file_name/var_name (defined by ENSO experts)
- Action: produce draft enso_ref_obs.py (Yann & Eric)
- Action: produce init_enso_metrics_obs.py (Jiwoo)

eguil · 2017-09-14T16:40:48Z

Meeting Jiwoo/Peter/Yann/Eric on Sept. 14 st:

action: Jiwoo to use new initEnsoMetrics.py and metricsCollectionsLib.py in driver all the way to run
action: Yann to organise local set up using same driver
action: Yann and Eric to add other metrics/collections
action: review metrics calculation structure to include difference as well as model + obs calculation (Yann)
document GitHub management for project #6

eguil · 2017-09-28T16:39:28Z

Meeting Jiwoo/Peter/Yann/Eric on Sept. 28th:

Yann presented workflow document stored on slack
organise exchange between PMP github and ENSO metrics gihub (PG & Jiwoo)
develop proposed ENSOCollectionsLib.py and ENSOMetricsLib.py (Yann)
init.py not needed (belongs to driver)
work on ENSO_metrics_driver.pyand associated my_param.py in dedicated branch (Jiwoo and PG)

eguil · 2017-10-12T16:59:07Z

Meeting Jiwoo/Yann/Charles/Eric on Oct 12th:

Charles describes new organisation on github towards creating conda packages (when the code is stabilized)
For now Jiwoo will work in the scripts/ directory (move parser.py back to scripts/)
Jiwoo presents new devs on driver
discussion how to define dates for observation and models (test on common/min length) - will live in ENSOCollectionsLib.py
discussion on re-organising the metrics collection dicts (Yann and Eric)
create ENSOcdmsToolsLib.py with basic routines used by ENSOMetrics.lib
next meeting Oct 26th

eguil · 2017-10-27T06:44:19Z

Meeting Jiwoo/Yann/Peter/Eric on Oct 27th:

Busan meeting summary
ENSO Research Foci Busan Oct 2017.pptx
IS-ENES3 discussions
Discussion on display methods of MME results (Jiwoo / Peter)
List of observations used for Obs4MIPs (Yann)
rest of actions from last time
next meeting Nov 2nd (Jiwoo and Yann)

eguil · 2017-11-10T17:14:06Z

Meeting Jiwoo/Yann/Peter/Eric on Nov 10th:

Yann shows new organisation of ENSOCollectionLib.py and other routines
TODO: PMPDriver update to new organisation [Jiwoo]
PG informs the group of precip work who could use the same framework
next meeting Nov 27th

Ongoing actions:

Document different display methods for MME results (Jiwoo / Peter) [example for ENSO metrics]
List of observations used for Obs4MIPs (Yann)
Regridding workflow

eguil · 2017-11-30T17:53:02Z

Meeting Jiwoo,Yann, Eric on Nov 30th:

Jiwoo updated driver to new organisation of metrics collection (and corrected some bugs)
ENSO RF telco on ENSO metrics def last week - see channel #enso_rf_discussion in slack
TODO (Yann): propose compute_collection() (which may include a loop oncompute_metric()):
- compute metric as model-obs distance for each metric for one model
- have to define this distance calculation for each metric
- the idea is to leave as little as possible to driver (only loop on models, path to files)
- propose normalisation options to plot all metrics on same figure in a later stage

Ongoing actions:

Document different display methods for MME results (Jiwoo / Peter) [example for ENSO metrics]
List of observations used for Obs4MIPs (Yann)
Regridding workflow

eguil · 2017-12-14T17:26:19Z

Meeting Jiwoo, Yann, Eric on Dec 14th:

Yann demonstrates the new compute_collection() and associated procedures
the term metric is now reserved for the distance between model and observation
Driver will use info from my_Param.py to create dicts loop for compute_collection()
new metrics to be implemented for next telco (TODO Yann)
Jiwoo update use of database interface
TODO: Jiwoo to adapt new version (commited now, comments needed)

Ongoing actions:

Document different display methods for MME results (Jiwoo / Peter) [example for ENSO metrics]
List of observations used for Obs4MIPs (Yann)
Regridding workflow

eguil · 2017-12-21T17:48:40Z

Meeting Jiwoo, Yann, Eric on Dec 21st:

Jiwoo's work last week:
- confirmed the revised code by Yann which works well with the PMP driver
- Build consistent XML files so I can easily apply all CMIP5 models to current version of code and parameter file
- Revisited the portrait plot to clean up code and prepare documenting
Yann's work:
- created a branch for new version (called yann for now but we should find a new name)
- further development of compute_collection() and associated tools (regrid, filter,...)
- new MC2
TODO [Jiwoo]: use compute_collection() from driver to remove loop on metrics - just keep loop on models
TODO [Eric]: start a shared document describing requirements for interface from this first experience

Ongoing actions:

Document different display methods for MME results (Jiwoo / Peter) [example for ENSO metrics]
List of observations used for Obs4MIPs (Yann)
Regridding workflow

eguil · 2018-01-10T17:53:28Z

Meeting Jiwoo, Yann, Peter, Eric on Jan 10th:

PG describes how an obs database is being developed in addition to obs4MIPs, using the same format (second bullet of ongoing actions)
PG suggest to use several regridding to assess impact (ESMF, regrid2,...)
EG suggest to not develop the interface in too much detail before we start interacting with ESMValTool.
Jiwoo completed pre-processor to ease use of driver
TODO [Jiwoo]: use compute_collection() from driver to remove loop on metrics - just keep loop on models
TODO [Eric]: start a shared document describing requirements for interface from this first experience
Yann developing MC2
Need to check all observations required are present

Ongoing actions:

Document different display methods for MME results (Jiwoo / Peter) [example for ENSO metrics]
List of observations used to be put in new ref obs database
Regridding workflow

eguil · 2018-01-26T17:57:49Z

Meeting Jiwoo, Yann, Peter, Eric on Jan 26th:

Yann and Peter: 2 normalization techniques illustrated in slack (vs. median or vs. ratio to obs)
- may want to separate different statistics in multi window plots with different color scales
PG: started to prepare observational database to follow obs4MIPs structure
Jiwoo: driver now calls compute_collection() - see metrics_collection slack channel and MC2 branch on github - required changes in obs database and internal organisation of driver
Yann: developed MC2 collection (from first proposal by CLIVAR group) and tested them
TODO [Yann]: explore how to override the collection default period per metric
TODO [Jiwoo]: consolidate driver code, try different regrid
TODO [Yann]: think about regrid specification (metric level, driver level, ?)
first meeting with ESMValTool next week (Feb 1st)

Ongoing actions:

TODO [Eric]: start a shared document describing requirements for interface from this first experience

eguil · 2018-02-02T18:06:56Z

Meeting Jiwoo, Yann, Peter, Eric on Feb 2nd:

summary of meeting with Veronika and Axel
idea to move pre-preprocessing to driver using specific encoding in ENSOCollectionsLibs.py
TODO [Yann+Jiwoo]: prepares a branch to test extracting pre-processing to driver
TODO [Jiwoo]: explore how to put portrait plot and provenance (per metrics) on a web/wiki
TODO [Jiwoo]: run MC2 to produce a portrait plot

Ongoing actions:

Eric: start a shared document describing requirements for interface from this first experience (in wiki to start with)
PG and JL to explore and document 1) regriding impact on results and 2) portrait plot normalization

eguil · 2018-02-15T18:03:16Z

Meeting Jiwoo, Yann, Peter, Eric on Feb 15th:

Yann shows the pre-process branch with just one metric
TODO [Jiwoo]: try again the MC2 branch to identify bug
Once this is working, then move to pre-process branch and update the rest of MC1 and MC2
Put pre-process branch back to master and work from there (create issues, related branches and pull back to master once issue is solved)

Ongoing actions:

Eric: start a shared document describing requirements for interface from this first experience (in wiki to start with)
PG and JL to explore and document 1) regriding impact on results and 2) portrait plot normalization

eguil · 2018-02-22T17:29:16Z

Meeting Jiwoo, Yann, Eric on Feb 22th

2 bugs identified (time checking, jason file issue wrt use of "mv2.multiply" vs. "*")
jason file made for the IPSL model / MC1
still an issue with one model
Once this is working, then move to pre-process branch and update the rest of MC1 and MC2, put pre-process branch back to master and work from there (create issues, related branches and pull back to master once issue is solved)
Jiwoo introduces Vega visualisation for jason files
Meeting with ESMValTool group see ESMValTool integration #14

Ongoing actions:

Eric: start a shared document describing requirements for interface from this first experience (in wiki to start with)
PG and JL to explore and document 1) regriding impact on results and 2) portrait plot normalization

yyplanton · 2018-03-08T17:56:32Z

Meeting (a quick hello from Eric) Jiwoo, Peter, Yann, on Mar 8th

uvcdat's cdutil.averager cannot handle irregular grids => we use areacello to average oceanic variables
for now Jiwoo added areacello to the xml file of oceanic variables but it should be in separate files
if we still have issues with averaging oceanic variables => use atmospheric ts (instead of tos) and tauu (instead of tauuo)

Ongoing actions:

Yann: merge MC2 branch to master
Yann: finalize preprocess branch
Jiwoo: run preprocess branch and take ownership of pre-processing functions (Yann will help to understand and develop them)
PG and JL to explore and document 1) regriding impact on results and 2) portrait plot normalization

eguil · 2018-03-20T16:55:41Z

Meeting Jiwoo, Peter, Yann, Eric on Mar 20th

MC1 and MC2 json file now working
portrait plot in the making - issue of structure (nested dict + model AND obs)
- leave models out that don't have the right data for now
work on identifying ESMValTool pre-processing
bring MC2 branch into master [DONE]
use preprocess branch to investigate interface for ESMValTool
keep separate until ESMValTool interface works and then merge in Master

Ongoing actions:

PG and JL to explore and document 1) regriding impact on results and 2) portrait plot normalization 3) verification suite using correlations

eguil · 2018-04-05T16:46:22Z

Meeting Jiwoo, Peter, Yann, Eric on April 5th:

MC2 merged in to master
Portrait plots provided by Jiwoo
relative errors shown wrt to mean error - ok to identify outliers
TODO [Jiwoo] explore absolute error view:
- parallel coordinates plot
- use of scale factor needed to show absolute error
- add metric numbers in boxes to combine relative and absolute

Book keeping:

use preprocess branch to investigate interface for ESMValTool
keep separate until ESMValTool interface works and then merge in Master

Ongoing actions:

PG and JL to explore and document 1) regriding impact on results and 2) portrait plot normalization 3) verification suite using correlations

eguil · 2018-05-03T16:37:56Z

Meeting Jiwoo, Peter, Yann, Eric on May 3rd:

new portrait plots with actual metrics value as number on slack
add unit [Jiwoo]
create legend (online or doc) [Yann]
rename MC2 into "ENSO perf"
to start interacting with Andrew and al:
- make version with atmos fields in new branch:atmos_regrid [Yann]
- try other regridding options - add option in current code to allow overwrite by driver [Yann]
- [add GFDL-CM2.1 (using atmos field is necessary) [Jiwoo]]
- change relative into absolute metric calculation [Yann]
think about adding other obs references (triangle)

Book keeping:

use preprocess branch to investigate interface for ESMValTool
keep separate until ESMValTool interface works and then merge in Master

Ongoing actions:

PG and JL to explore and document 1) regriding impact on results and 2) verification suite using correlations

eguil · 2018-05-17T16:38:28Z

Meeting Jiwoo, Peter, Yann, Eric on May 17th:

work on regrid done by Yann ( to be commited soon)
move from matplotlib to VCS for graphics (better for interactive plots) done by Jiwoo
add units [Jiwoo]
create legend (online or doc) [Yann]
rename MC2 into "ENSO perf"
work on regrid (e.g. ts vs. tos): add option to have it per variable
look into Yann being able to rerun the portrait plot by himself

Book keeping:

use preprocess branch to investigate interface for ESMValTool
keep separate until ESMValTool interface works and then merge in Master

Ongoing actions:

PG and JL to explore and document 1) regriding impact on results and 2) verification suite using correlations
think about adding other obs references (triangle)

eguil · 2018-05-29T16:32:20Z

Meeting Jiwoo, Peter, Yann, Eric on May 29th:

regrid work by Yann in branch regrid:
- added option user_regridding=<option> in Compute_Collection call in driver
- documented in README_regridding in scripts
- works per collection or per metrics
- tested in a few cases
- next step: Jiwoo to test this
set up call with Andrew weeks 3 and 4 of June
- units
- redo a few portrait plots (different combinations of color/numbers)
- documentation
Next major milestone: portrait plots of ENSO perf with validated science and cleaned up codes

Book keeping:

use preprocess branch to investigate interface for ESMValTool
keep separate until ESMValTool interface works and then merge in Master
look into Yann being able to rerun the portrait plot by himself: wait for clean up of code for now

Ongoing actions:

PG and JL to explore and document 1) regriding impact on results and 2) verification suite using correlations
think about adding other obs references (triangle)
further organise provenance (both technical and scientific)

eguil · 2018-07-16T16:54:50Z

Meeting Jiwoo, Peter, Yann, Eric on July 16th:

json output reviewed to fit wider framework - update dict{} structure to fit this requirement (work around currently exists)
technical need to create a tag (e.g. v0.1) to create conda version to start integration in PMP
discuss science needs from telco with Andrew:
- more portrait plot with absolute/relative metrics
  - interesting plot on slack (june 28th) mixing metric and relative MME color
  - TODO: [Jiwoo] have 2 portraits plots (RMSE & non RMSE)
- // coord plots
- dive down plots
- science prioritisation (correlations between metrics)
new version of code pushed by Yann Friday
- some dive down diags included (values used to compute the metric)
- [Jiwoo] to start producing a plot for dive down
MC news:
- [Yann] start work on ENSO teleconnections

Ongoing actions:

PG and JL to explore and document 1) regriding impact on results and 2) verification suite using correlations
think about adding other obs references (triangle)
further organise provenance (both technical and scientific)

yyplanton · 2018-07-26T16:47:44Z

Meeting Jiwoo, Yann on July 26th:

technical need to create a tag (e.g. v0.1): done. Jiwoo explained to Yann how to do it
// coord plots created by Jiwoo (with or without lines): very nice to see the spread among models
- it would be nice to be able to select several models for the plot and not always all of them (interactive plots)
- is there a way to have the full markers for the first un last metrics of the plot? (and not half of it as it is now)
discuss science needs from telco with Andrew:
- more portrait plot with absolute/relative metrics
  - interesting plot on slack (june 28th) mixing metric and relative MME color
  - TODO: [Jiwoo] have 2 portraits plots (RMSE & non RMSE)
- interesting // coord plots on slack (july 24th)
- TODO: [Jiwoo] to start producing a plot for dive down (using the last version of the atmos_regrid branch)
- science prioritisation (correlations between metrics)
MC teleconnections:
- [Yann] first two metrics will be available quickly (still some science to discuss with Scott Power)
- [Yann] these metrics will be used to test the json output reviewed to fit wider framework - update dict{} structure to fit this requirement

Ongoing actions:

PG and JL to explore and document 1) regriding impact on results and 2) verification suite using correlations
think about adding other obs references (triangle)
further organise provenance (both technical and scientific)

eguil · 2018-08-16T16:53:15Z

Meeting Jiwoo, Peter, Yann, Eric on Aug. 16th:

discussion on parallel plots:
- very nice complement to portrait plot
- issues with taux (2 models) and sst (?)
version testing ready - will be included in PMP release for testing (hidden)
Yann: MC3 on ENSO teleconnection: branch Test_MC3
- science discussion with Scott P on going
format of the dict for the json format
- same depth level currently has mix of metrics (values) and meta-data (strings)
- fix structure to have only one type: how ?
  - work on package side or work on server side ?
  - Yann & Jiwoo to test: have 2 dict (current and data only) embedded into one
- Jiwoo and Peter to create an issue to document that, Yann uses the branch
discuss science needs from telco with Andrew:
- more portrait plot with absolute/relative metrics
  - TODO: [Jiwoo] have 2 portraits plots (RMSE & non RMSE)
- interesting // coord plots on slack (july 24th)
- TODO: [Jiwoo] to start producing a plot for dive down (using the last version of the atmos_regrid branch)
- science prioritisation (correlations between metrics)
MC teleconnections:
- [Yann] first two metrics will be available quickly (still some science to discuss with Scott Power)
- [Yann] these metrics will be used to test the json output reviewed to fit wider framework - update dict{} structure to fit this requirement

Ongoing actions:

PG and JL to explore and document 1) regriding impact on results and 2) verification suite using correlations
think about adding other obs references (triangle)
further organise provenance (both technical and scientific)

eguil · 2018-09-13T16:42:46Z

Meeting Jiwoo, Yann, Eric on Sept. 13th:

ENSO RF discussion two days ago
- summary of progress presented - people are getting excited by the results
- MC3 / ENSO teleconnections discussed
json format and dive down diags
- branch to test ideas
- pb with dive down diags (too long a file)
- Start with option for dive down -> creates a new json file
- but netCDF will eventually be required
- Yann will commit a version with new json structure (test by Jiwoo and bring back to trunk)
display topics
- TODO: [Jiwoo] have 2 portraits plots (RMSE & non RMSE)
- TODO: [jiwoo] // plot: issues with taux (2 models) and sst (clustering ?)
- Jiwoo demonstrates an interactive // plot - link on slack
- have science ready plots for ENSO meeting (Ecuador Oct 12-15th)

See above for other on-going actions

eguil · 2018-09-28T17:00:24Z

Meeting Jiwoo, Yann, Eric on Sept. 28th:

review tickets/issues
review branches
- atmos_regrid can go
- merge needed for json_structure - review pull request
correcting data
- taux: reversed - correct data rather than the code (driver issue)
- sst: using ocean-only points - ENSO metrics package calculation to integrate this
- currently debugging on test_MC3 branch (use one/two models for now)
Guayaquil material
- 2 portrait plots
- // coord. plot fixed
- // coord. interactive

eguil · 2018-10-08T16:33:04Z

Meeting Jiwoo, Yann, Eric on Oct. 8th:

material for Guayaquil todo before end of week:
- // plot fixed
- portrait non-RMSE (MC1)
- still an issue with some incoherent data/LS mask (driver issue ? e.g ACCESS)
- taux issue:
  - fixed in data but not yet in portrait - either remove or fix
Yann still investigating how issue was fixed (maybe interpolation and masking?)
branches reviews
- Yann still investigating so merge will wait

eguil · 2018-11-15T17:37:01Z

Meeting Jiwoo, Yann, Peter, Eric on Nov. 15th:

Meetings
- ENSO Guayaquil - clear support
- ENSO CLIVAR work
  - work on MC1 (5 new metrics) and MC2 (more changes coming up) [TODO: Yann]
  - suggestions to add options in // plot - keep track of wish list in new issue [TODO:Yann]
  - [TODO:Yann]: redo correlation in a more readable way and put on wiki
- CMIP6 Barcelona abstract < 15 Dec
  - ENSO metrics
- DOE meeting
  - overview of PMP and ENSO metrics - lot's of interest (other metrics such as MJO)
  - provide current work to demonstration site
Dive down diags
- stored in other json file - ok for 1D, more complex for 2D (use netCDF with option)
- naming of file: file of root sent by driver and metric name appended
MC ENSO perform two portrait plots RMSE, non-RMSE - add as property in output or rather add description in metric collection file

eguil · 2018-11-28T17:15:30Z

Meeting Jiwoo, Yann, Peter, Eric on Nov. 28th:

option to save netCDF ready in computeMetrics file
ENSO CLIVAR work
- work on MC1 (5 new metrics) and MC2 (more changes coming up) [TODO: Yann]
- suggestions to add options in // plot - keep track of wish list in new issue [TODO:Yann]
- [TODO:Yann]: redo inter-metrics correlation in a more readable way and put on wiki
- add references to wiki metrics info
- DONE: MC ENSO perform two portrait plots RMSE, non-RMSE - add description in metric collection file
CMIP6 Barcelona abstract < 15 Dec
- ENSO metrics

yyplanton · 2018-12-19T17:49:48Z

Meeting Jiwoo, Yann on Dec. 19th:

[TODO: Yann] option to save more information in netCDF
ENSO CLIVAR work
- [TODO: Yann] work on ENSO_perf (one new metric has still to be coded: spectrum) and ENSO_tel (more changes coming up)
- suggestions to add options in // plot - keep track of wish list in new issue [TODO:Yann]
- [TODO:Yann]: redo inter-metrics correlation in a more readable way and put on wiki
- add references to wiki metrics info
- DONE: MC ENSO perform two portrait plots RMSE, non-RMSE - add description in metric collection file
- [TODO:Yann]: give Jiwoo me code to plot dive down diagnostics

lee1043 · 2020-07-13T19:51:06Z

Closing this issue for house cleaning

eguil mentioned this issue Jul 27, 2017

Merge with PMP repo #3

Closed

eguil changed the title ~~Work plan summer/fall 2017~~ Work plan PMP 2017/2018 Feb 22, 2018

eguil closed this as completed May 3, 2018

eguil reopened this May 3, 2018

lee1043 closed this as completed Jul 13, 2020

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Work plan PMP 2017/2018 #4

Work plan PMP 2017/2018 #4

eguil commented Jul 27, 2017 •

edited

Loading

eguil commented Aug 31, 2017 •

edited

Loading

eguil commented Sep 14, 2017 •

edited

Loading

eguil commented Sep 28, 2017 •

edited

Loading

eguil commented Oct 12, 2017 •

edited

Loading

eguil commented Oct 27, 2017

eguil commented Nov 10, 2017 •

edited

Loading

eguil commented Nov 30, 2017 •

edited

Loading

eguil commented Dec 14, 2017

eguil commented Dec 21, 2017 •

edited

Loading

eguil commented Jan 10, 2018 •

edited

Loading

eguil commented Jan 26, 2018

eguil commented Feb 2, 2018 •

edited

Loading

eguil commented Feb 15, 2018

eguil commented Feb 22, 2018 •

edited

Loading

yyplanton commented Mar 8, 2018

eguil commented Mar 20, 2018 •

edited

Loading

eguil commented Apr 5, 2018 •

edited

Loading

eguil commented May 3, 2018 •

edited

Loading

eguil commented May 17, 2018

eguil commented May 29, 2018 •

edited

Loading

eguil commented Jul 16, 2018 •

edited

Loading

yyplanton commented Jul 26, 2018 •

edited

Loading

eguil commented Aug 16, 2018 •

edited

Loading

eguil commented Sep 13, 2018 •

edited

Loading

eguil commented Sep 28, 2018

eguil commented Oct 8, 2018

eguil commented Nov 15, 2018 •

edited

Loading

eguil commented Nov 28, 2018 •

edited

Loading

yyplanton commented Dec 19, 2018 •

edited

Loading

lee1043 commented Jul 13, 2020

Work plan PMP 2017/2018 #4

Work plan PMP 2017/2018 #4

Comments

eguil commented Jul 27, 2017 • edited Loading

eguil commented Aug 31, 2017 • edited Loading

eguil commented Sep 14, 2017 • edited Loading

eguil commented Sep 28, 2017 • edited Loading

eguil commented Oct 12, 2017 • edited Loading

eguil commented Oct 27, 2017

eguil commented Nov 10, 2017 • edited Loading

eguil commented Nov 30, 2017 • edited Loading

eguil commented Dec 14, 2017

eguil commented Dec 21, 2017 • edited Loading

eguil commented Jan 10, 2018 • edited Loading

eguil commented Jan 26, 2018

eguil commented Feb 2, 2018 • edited Loading

eguil commented Feb 15, 2018

eguil commented Feb 22, 2018 • edited Loading

yyplanton commented Mar 8, 2018

eguil commented Mar 20, 2018 • edited Loading

eguil commented Apr 5, 2018 • edited Loading

eguil commented May 3, 2018 • edited Loading

eguil commented May 17, 2018

eguil commented May 29, 2018 • edited Loading

eguil commented Jul 16, 2018 • edited Loading

yyplanton commented Jul 26, 2018 • edited Loading

eguil commented Aug 16, 2018 • edited Loading

eguil commented Sep 13, 2018 • edited Loading

eguil commented Sep 28, 2018

eguil commented Oct 8, 2018

eguil commented Nov 15, 2018 • edited Loading

eguil commented Nov 28, 2018 • edited Loading

yyplanton commented Dec 19, 2018 • edited Loading

lee1043 commented Jul 13, 2020

eguil commented Jul 27, 2017 •

edited

Loading

eguil commented Aug 31, 2017 •

edited

Loading

eguil commented Sep 14, 2017 •

edited

Loading

eguil commented Sep 28, 2017 •

edited

Loading

eguil commented Oct 12, 2017 •

edited

Loading

eguil commented Nov 10, 2017 •

edited

Loading

eguil commented Nov 30, 2017 •

edited

Loading

eguil commented Dec 21, 2017 •

edited

Loading

eguil commented Jan 10, 2018 •

edited

Loading

eguil commented Feb 2, 2018 •

edited

Loading

eguil commented Feb 22, 2018 •

edited

Loading

eguil commented Mar 20, 2018 •

edited

Loading

eguil commented Apr 5, 2018 •

edited

Loading

eguil commented May 3, 2018 •

edited

Loading

eguil commented May 29, 2018 •

edited

Loading

eguil commented Jul 16, 2018 •

edited

Loading

yyplanton commented Jul 26, 2018 •

edited

Loading

eguil commented Aug 16, 2018 •

edited

Loading

eguil commented Sep 13, 2018 •

edited

Loading

eguil commented Nov 15, 2018 •

edited

Loading

eguil commented Nov 28, 2018 •

edited

Loading

yyplanton commented Dec 19, 2018 •

edited

Loading