Add attr to the dataframe #129

xiki-tempula · 2021-05-15T14:08:32Z

codecov · 2021-05-16T21:04:59Z

Codecov Report

Merging #129 (06bda30) into master (ea0cb00) will increase coverage by 0.21%.
The diff coverage is 100.00%.

@@            Coverage Diff             @@
##           master     #129      +/-   ##
==========================================
+ Coverage   97.48%   97.70%   +0.21%     
==========================================
  Files          17       20       +3     
  Lines         996     1091      +95     
  Branches      220      230      +10     
==========================================
+ Hits          971     1066      +95     
  Misses          5        5              
  Partials       20       20

Impacted Files	Coverage Δ
src/alchemlyb/visualisation/convergence.py	`100.00% <ø> (ø)`
src/alchemlyb/__init__.py	`100.00% <100.00%> (ø)`
src/alchemlyb/estimators/bar_.py	`100.00% <100.00%> (ø)`
src/alchemlyb/estimators/mbar_.py	`96.55% <100.00%> (+0.25%)`	⬆️
src/alchemlyb/estimators/ti_.py	`100.00% <100.00%> (ø)`
src/alchemlyb/parsing/__init__.py	`100.00% <100.00%> (ø)`
src/alchemlyb/parsing/amber.py	`97.48% <100.00%> (+0.04%)`	⬆️
src/alchemlyb/parsing/gmx.py	`97.50% <100.00%> (+0.06%)`	⬆️
src/alchemlyb/parsing/gomc.py	`92.23% <100.00%> (+0.31%)`	⬆️
src/alchemlyb/parsing/namd.py	`100.00% <100.00%> (ø)`
... and 7 more

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update ea0cb00...06bda30. Read the comment docs.

xiki-tempula · 2021-05-21T07:55:57Z

@orbeckst Sorry for disturbing you. I wonder if you mind have a look and give some advice, please? Thank you.

I have added the deprecation of py2 and py356 into this pr for the test to pass but they could be moved to another PR.

orbeckst

Initial comments

make 2.7/3.5 removal separate PR
need to keep 3.6 around (why does pd not publish packages for 3.6 ???????) so we need to work-around

orbeckst · 2021-05-28T18:50:18Z

.travis.yml

@@ -7,9 +7,6 @@ env:
    - MPLBACKEND=agg

 python:
-   - "2.7"
-   - "3.5"
-   - "3.6"


need to keep 3.6 for now, see #130 , don't do the removal here

orbeckst · 2021-05-28T18:50:58Z

setup.py

-        'Programming Language :: Python :: 2.7',
-        'Programming Language :: Python :: 3',
-        'Programming Language :: Python :: 3.5',
-        'Programming Language :: Python :: 3.6',


keep 3.6 for now; don't do the removal here

orbeckst · 2021-05-28T19:00:30Z

src/alchemlyb/estimators/mbar_.py

@@ -92,7 +92,10 @@ def fit(self, u_nk):
                              columns=self.states_,
                              index=self.states_) for i in out]

-        (self.delta_f_, self.d_delta_f_, self.theta_) = attrs 
+        (self.delta_f_, self.d_delta_f_, self.theta_) = attrs


Maybe we want to rename those attrs to something else to avoid confusion. free_energy_differences ?

This could be something that could be improved.
but delta_f_, d_delta_f_ are used in all estimators and in the pymbar module.

I meant the use of the name attrs on the RHS. Now that we are using df.attrs, it is very confusing because these free energy diffs aren't the attrs that we are using for units.

orbeckst · 2021-05-28T19:18:25Z

src/alchemlyb/parsing/gmx.py

+    u_k.attrs['temperature'] = T
+    u_k.attrs['energy_unit'] = 'kT'
+


We could wrap attribute setting so that we can use old pandas.

# remove once we can require pandas >= 1.2 (after Python 3. if pd.__version__ >= "1.2": # do this version check properly def pd_setattrs(df, data): df.attrs = data else: def pd_setattrs(df, data): if not isinstance(data, dict): raise TypeError with warnings.catch_warnings(): warnings.simplefilter("ignore") setattr(df, 'attrs', data)

and then use pd_setattrs() everywhere where you set the attributes for the first time.

This could live in util or somewhere similar.

@orbeckst Sorry, it is a bit unclear to me of what this function is trying to achieve. There should be no problem if one were to only use the alchemlyb functions when pandas < 1.2, but if one were to manipulate the data frame by themselves, pandas < 1.2 would then drop the attrs.

When I tried to set df.attrs, pandas complained that I was trying to add a column but set the attribute nevertheless. My idea was to just filter the warnings.

It was not clear to me what the problem with attrs for 1.0 ≤ pandas < 1.2 was: Are you saying that only for ≥ 1.2, slicing a df will retain the attrs on the sliced copy?

@orbeckst

When I tried to set df.attrs, pandas complained that I was trying to add a column but set the attribute nevertheless. My idea was to just filter the warnings.

For pandas == 1.0.5,

>>> d = {'col1': [1, 2], 'col2': [3, 4]} >>> df = pd.DataFrame(data=d) >>> df.attrs {} >>> df.attrs = {1:1}

yields no warning for me.

It was not clear to me what the problem with attrs for 1.0 ≤ pandas < 1.2 was: Are you saying that only for ≥ 1.2, slicing a df will retain the attrs on the sliced copy?

Yes.

for pandas >=1.0 and < 1.2

>>> import pandas as pd >>> d = {'col1': [1, 2], 'col2': [3, 4]} >>> df = pd.DataFrame(data=d) >>> df.attrs = {1:1} >>> df.attrs {1: 1} >>> df['col1'].attrs {}

orbeckst · 2021-05-31T17:39:31Z

@xiki-tempula I am a bit busy over the next few days, sorry if I don't get to reviewing anything. Please feel free to ping me around Wed/Thu again. Thank you for your understanding & patience. I really appreciate the work you're putting in here.

orbeckst

Looking already quite good. I had various comments and one important additional request: We need to document this change in behavior under https://alchemlyb.readthedocs.io/en/stable/parsing.html#standard-forms-of-raw-data so please add text that explains how we annotate units and what users have to do (namely, use pass_attrs.

Furthermore, add documentation about the new postprocessing module to the docs.

Update the visualization docs because now units are used for conversion.

Update CHANGES (these are big important ones).

We might need another round of reviews but this is already looking really nice. Great work @xiki-tempula !

src/alchemlyb/estimators/bar_.py

src/alchemlyb/estimators/mbar_.py

src/alchemlyb/parsing/gmx.py

orbeckst · 2021-06-03T18:26:34Z

src/alchemlyb/postprocessors/units.py

+    assert 'temperature' in df.attrs, 'Attribute temperature not found in the' \
+                                      ' input Dataframe.'
+    assert 'energy_unit' in df.attrs, 'Attribute energy_unit not found in the' \
+                                      ' input Dataframe.'


Replace assert with raising a TypeError, saying that we need a df with attrs according to https://alchemlyb.readthedocs.io/en/stable/parsing.html#standard-forms-of-raw-data . Ultimately, the dfs are user input so we have to check them properly.

(assert can be optimized away and should not be used to check user input.)

src/alchemlyb/postprocessors/units.py

orbeckst · 2021-06-03T18:42:16Z

src/alchemlyb/tests/test_units.py

+        assert new_dhdl.attrs['temperature'] == 310
+        assert new_dhdl.attrs['energy_unit'] == 'kT'
+
+class Test_estimator():


add as TestUnits to the test_*_estimators.py so that in the future it is easier to add tests when anything changes in estimators.

orbeckst · 2021-06-03T18:42:39Z

src/alchemlyb/tests/test_units.py

+        assert mbar.d_delta_f_.attrs['temperature'] == 300
+        assert mbar.d_delta_f_.attrs['energy_unit'] == 'kT'
+
+class Test_visualisation():


add as TestUnits to the test_visualisation.py so that in the future it is easier to add tests when anything changes in viz.

orbeckst · 2021-06-03T18:44:35Z

src/alchemlyb/visualisation/dF_state.py

+                    dF.append(to_kcalmol(dhdl.delta_f_).iloc[i, i + 1])
+                    error.append(to_kcalmol(dhdl.d_delta_f_).iloc[i, i + 1])
+                else:
+                    raise NameError('energy_unit {} can only be kT, kJ/mol '


ValueError – but isn't that already being raised with the conversion functions?

If you use the dispatch table approach, it's already included.

orbeckst · 2021-06-03T18:49:56Z

src/alchemlyb/visualisation/dF_state.py

+                if units == 'kT':
+                    dF.append(to_kT(dhdl.delta_f_).iloc[i, i + 1])
+                    error.append(to_kT(dhdl.d_delta_f_).iloc[i, i + 1])
+                elif units == 'kJ/mol':
+                    dF.append(to_kJmol(dhdl.delta_f_).iloc[i, i + 1])
+                    error.append(to_kJmol(dhdl.d_delta_f_).iloc[i, i + 1])
+                elif units == 'kcal/mol':
+                    dF.append(to_kcalmol(dhdl.delta_f_).iloc[i, i + 1])
+                    error.append(to_kcalmol(dhdl.d_delta_f_).iloc[i, i + 1])


Too much code duplications. Use a dispatch table

_converters = {'kT': to_kT, 'kJ/mol': to_kJmol, 'kcal/mol': to_kcalmol} try: convert = _converters[units] except KeyError: raise ValueError(f"Energy unit {units} is not supported, choose one of {list(_converters.keys())}") dF.append(convert(dhdl.delta_f_).iloc[i, i + 1]) error.append(convert(dhdl.d_delta_f_).iloc[i, i + 1])

You could put the dispatch table into units and import it.

orbeckst · 2021-06-03T18:51:05Z

src/alchemlyb/visualisation/ti_dhdl.py

+    for dhdl in dhdl_list:
+        if units == 'kT':
+            new_unit.append(to_kT(dhdl))
+        elif units == 'kJ/mol':
+            new_unit.append(to_kJmol(dhdl))
+        elif units == 'kcal/mol':
+            new_unit.append(to_kcalmol(dhdl))
+        else:
+            raise NameError('energy_unit {} can only be kT, kJ/mol or ' \
+                            'kcal/mol.'.format(units))


use dispatch table

xiki-tempula · 2021-06-12T14:47:22Z

@orbeckst Thanks for the review and the advice. I think I have addressed all the changes. Would you mind have a look, please? Thank you.

orbeckst · 2021-06-14T17:29:50Z

I won't be able to review until middle of the week, sorry. Thanks for your patience.

orbeckst · 2021-06-17T15:23:29Z

Can you please rebase against master so that the new CI runs? Thanks.

xiki-tempula · 2021-06-17T15:45:26Z

@orbeckst Thanks for the advice. I have merged the master to this branch.

orbeckst

looks pretty good, some minor comments.

I didn't manage to look at everything in as much detail as I wanted, If I see more I add it.

orbeckst · 2021-06-18T00:45:05Z

docs/parsing.rst

+    >>> dhdl_coul = alchemlyb.concat([extract_dHdl(xvg, T=300) for xvg in
+    dataset['Coulomb']])


Don't line break

orbeckst · 2021-06-18T00:45:30Z

docs/parsing.rst

+
+The metadata (such as the unit of the energy and temperature) are stored in 
+:attr:`pandas.DataFrame.attrs`, a :class:`dict`. Functions in ``alchemlyb`` are
+aware


reflow (aware is alone on its line)

src/alchemlyb/__init__.py

orbeckst · 2021-06-18T00:52:42Z

src/alchemlyb/__init__.py

+    '''Concatenate pandas objects along a particular axis with optional set
+    logic along the other axes. If all pandas objects have the same attrs
+    attribute, the new pandas objects would have this attrs attribute. A
+    ValueError would be raised if any pandas object has a different attrs.


Add a Parameters section (see numpydoc)

orbeckst · 2021-06-18T00:53:14Z

src/alchemlyb/__init__.py

+    ValueError
+        If not all pandas objects have the same attrs.
+
+


Add a See Also section for pandas.concat.

Check if this renders correctly. If we don't have pandas in the intersphinx mapping, add it to sphinx conf.py.

Also make sure that there are at least two empty lines before the version added.

Yes, it is rendered correctly.

orbeckst · 2021-06-18T01:02:07Z

src/alchemlyb/postprocessors/units.py

+kJ2kcal = 1 / calorie
+R_kJmol = R / 1000


add doc as in

Suggested change

kJ2kcal = 1 / calorie

R_kJmol = R / 1000

#: conversion factor from kJ to kcal, based on :data:`scipy.constants.calorie`

kJ2kcal = 1 / calorie

#: Gas constant :math:`R` in kJ/(mol K), based on :data:`scipy.constants.R`

R_kJmol = R / 1000

I like your approach here.

Also add these constants to the documentation somewhere

.. autodata:: kJ2kcal .. autodata:: R_kJmol

docs/postprocessors/alchemlyb.postprocessors.units.rst

orbeckst · 2021-06-18T01:09:39Z

src/alchemlyb/postprocessors/units.py

+    kt_df.attrs['energy_unit'] = 'kJ/mol'
+    return kt_df
+
+def get_unit_converter(units):


make it appear in docs

orbeckst · 2021-06-18T01:09:50Z

src/alchemlyb/postprocessors/units.py

+    Returns
+    -------
+    func
+        converter.


no period; maybe link to the docs that list the possible converter functions

add a versionadded

… attr

xiki-tempula · 2021-06-19T10:36:55Z

@orbeckst Thanks for the review. I have updated the docs and checked it.

orbeckst

Second part of review.

Major question: Why do the GMX tests now produce different answers? This is a serious change and we need to be very clear about what changed, why, and how much of a difference it makes — if the change is necessary and not a bug.
There are tests where attr is explicitly assigned after use of extract_dHdl and alchemlyb.concat() and I don't understand why: I thought that alchemlyb will seamlessly handle units when using only alchemlyb functions.
Mostly minor doc issues.
Add postprocessing to docs/api_principles.rst

orbeckst · 2021-06-21T07:06:58Z

src/alchemlyb/__init__.py

+    ValueError
+        If not all pandas objects have the same attrs.
+
+


Check if this renders correctly. If we don't have pandas in the intersphinx mapping, add it to sphinx conf.py.

Also make sure that there are at least two empty lines before the version added.

orbeckst · 2021-06-21T07:09:51Z

CHANGES


 Deprecations

 Fixes

 Changes
+  - alchemlyb.concat added to replace pd.concat (PR #129).
+  - postprocessors.units module for unit conversion (PR #129).


Add under enhancements. (The "Changes" parts are listed separately, see other comments.)

Reference #125 (generally, the issue is more important than the PR, you can include both if you like, but definitely reference the issue)

orbeckst · 2021-06-21T07:10:14Z

CHANGES

@@ -18,12 +18,18 @@ The rules for this file:
  * 0.5.0

 Enhancements
+  - Visualisation module will change the data according to input unit (PR


Reference #125 (generally, the issue is more important than the PR, you can include both if you like, but definitely reference the issue)

orbeckst · 2021-06-21T07:11:07Z

CHANGES

@@ -18,12 +18,18 @@ The rules for this file:
  * 0.5.0

 Enhancements
+  - Visualisation module will change the data according to input unit (PR
+  #129).
+  - The parser output dataframe with metadata (PR #129).


I would list this under Changes because it changes our data format.

orbeckst · 2021-06-21T07:12:57Z

docs/estimators-fep.rst

-    >>> u_nk_coul = pd.concat([extract_u_nk(xvg, T=300) for xvg in bz['Coulomb']])
-    >>> u_nk_vdw = pd.concat([extract_u_nk(xvg, T=300) for xvg in bz['VDW']])
+    >>> u_nk_coul = alchemlyb.concat([extract_u_nk(xvg, T=300) for xvg in bz['Coulomb']])
+    >>> u_nk_vdw = alchemlyb.concat([extract_u_nk(xvg, T=300) for xvg in bz['VDW']])


I assume you did a git grep concat and looked at every occurrence in the code?

Yes, I did it within the pycharm IDE.

orbeckst · 2021-06-21T07:26:40Z

src/alchemlyb/tests/test_visualisation.py

+        bz = load_benzene().data
+        dHdl_coul = alchemlyb.concat(
+            [extract_dHdl(xvg, T=300) for xvg in bz['Coulomb']])
+        dHdl_coul.attrs = extract_dHdl(load_benzene().data['Coulomb'][0], T=300).attrs


Why attr assignment?

orbeckst · 2021-06-21T07:26:46Z

src/alchemlyb/tests/test_visualisation.py

+
+        u_nk_coul = alchemlyb.concat(
+            [extract_u_nk(xvg, T=300) for xvg in bz['Coulomb']])
+        u_nk_coul.attrs = extract_dHdl(load_benzene().data['Coulomb'][0], T=300).attrs


Why attr assignment?

orbeckst · 2021-06-21T07:29:33Z

src/alchemlyb/visualisation/convergence.py

    The units variable is for labelling only. Changing it doesn't change the
-    unit of the underlying variable, which is in the unit of :math:`kT`.
+    unit of the underlying variable.


Does plot_convergence() not do unit conversion, even though the other plotting functions do?

Can you change it, not the least the CHANGES says that this is now working for all visualization functions.

Or remove units and take the label from the df attr — but having the same behavior as the other plotting functions would be much better and more consistent.

@orbeckst There is currently no alchemlyb function that will generate the input for plot_convergence.
The user needs to generate the input file, which could be a simple python list.

src/alchemlyb/visualisation/dF_state.py

src/alchemlyb/visualisation/ti_dhdl.py

xiki-tempula · 2021-06-21T09:19:23Z

@orbeckst Thank you for the review. I will do the changes when I got time.

Major question: Why do the GMX tests now produce different answers? This is a serious change and we need to be very clear about what changed, why, and how much of a difference it makes — if the change is necessary and not a bug.

The original parser uses the gromacs constant https://manual.gromacs.org/documentation/2019/reference-manual/definitions.html

R | gas constant | 8.3144621×10−3 kJ mol−1 K−1

The new R uses the scipy constant.

>>> from scipy.constants import R
>>> R
8.314462618

They differ by (8.3144621-8.314462618)/8.3144621 = -6.230108382200071e-08.

There are tests where attr is explicitly assigned after use of extract_dHdl and alchemlyb.concat() and I don't understand why: I thought that alchemlyb will seamlessly handle units when using only alchemlyb functions.

Sorry, I initially did this when we only have pd.concat.
Then I felt it might be more useful to have alchemlyb.concat() instead of doing this everywhere, so I made it and change pd.concat to alchemlyb.concat. But I forget to remove this.

orbeckst · 2021-06-23T22:27:44Z

Thanks for the explanation regarding R. Given that we can consistently convert back to the actual kJ/mol units that GROMACS provides, the change in the reference values in the tests is ok.

Let me know when the PR is ready for final review.

orbeckst · 2021-06-23T22:29:58Z

We will need to add a note to CHANGES and a Warning to the gmx parser page, though, because the results in kT that people produced previously will now be slightly off.

xiki-tempula · 2021-06-26T10:24:42Z

@orbeckst Thanks for the review. I have addressed all the changes.
However, there seems to be a problem. I think I have configured the sphinx/conf.py correctly.
But :attr:pandas.DataFrame.attrs cannot be reference correctly.

orbeckst

Nice work, big step forward.

I have added minor corrections, which I will add in batch.

CHANGES

src/alchemlyb/parsing/amber.py

src/alchemlyb/parsing/gmx.py

src/alchemlyb/parsing/gomc.py

src/alchemlyb/parsing/namd.py

src/alchemlyb/parsing/gmx.py

orbeckst · 2021-06-28T19:59:32Z

docs/conf.py

-                       'alchemtest': ('https://alchemtest.readthedocs.io/en/latest/', None)}
+                       'alchemtest': (
+                           'https://alchemtest.readthedocs.io/en/latest/', None),
+                       'scipy': ('http://docs.scipy.org/doc/scipy/reference/', None)}


Suggested change

'scipy': ('http://docs.scipy.org/doc/scipy/reference/', None)}

'scipy': ('https://docs.scipy.org/doc/scipy/reference/', None)}

- explain change in GROMACS output to ~1e-7 - minor doc fixes

xiki-tempula added 2 commits May 15, 2021 15:07

add attr

9ce3301

Update setup.py

54439eb

xiki-tempula mentioned this pull request May 15, 2021

A unified way of unit conversion #125

Closed

xiki-tempula added 5 commits May 15, 2021 16:04

fix test

875b691

Update setup.py

cfef17d

Update setup.py

8451ca8

Update setup.py

34eb679

deprecate py2

9c7579b

xiki-tempula added 2 commits May 16, 2021 23:15

bump test

77111cb

Update util.py

5ec986a

orbeckst reviewed May 28, 2021

View reviewed changes

fix test

1913e86

xiki-tempula mentioned this pull request May 28, 2021

drop support for py2 and py35 bump pandas support to 1.0 #131

Merged

xiki-tempula added 2 commits May 28, 2021 20:53

add support for py36

cab1317

Update mbar_.py

78f8226

Merge branch 'master' into attr

560d700

orbeckst requested changes Jun 3, 2021

View reviewed changes

orbeckst self-assigned this Jun 3, 2021

xiki-tempula and others added 7 commits June 5, 2021 10:50

change parser

ce47b0d

change idea

8841f4b

bug fix

4605305

fix bug

920fba0

Update test_units.py

2b3f668

refactor

9977c26

bump test

1e31968

xiki-tempula marked this pull request as draft June 5, 2021 18:27

do the concat

ab06a71

xiki-tempula added 2 commits June 12, 2021 17:14

fix doc

07a034d

Update units.py

b4fd15a

xiki-tempula added 2 commits June 17, 2021 16:40

Delete .travis.yml

6592e6e

Merge branch 'master' into attr

583a198

orbeckst requested changes Jun 18, 2021

View reviewed changes

xiki-tempula added 4 commits June 19, 2021 10:53

fix docs

a1e9019

Merge branch 'attr' of https://github.com/xiki-tempula/alchemlyb into…

68184c3

… attr

update docs

9ed4917

fix docs

0af71be

Update units.py

ef05f52

orbeckst requested changes Jun 21, 2021

View reviewed changes

xiki-tempula marked this pull request as draft June 21, 2021 09:19

xiki-tempula added 3 commits June 26, 2021 10:50

change

551c081

change doc

db1e801

Update conf.py

da95d7d

xiki-tempula marked this pull request as ready for review June 26, 2021 10:24

update docs

035d088

orbeckst approved these changes Jun 28, 2021

View reviewed changes

doc updates

06bda30

- explain change in GROMACS output to ~1e-7 - minor doc fixes

xiki-tempula merged commit dce2893 into alchemistry:master Jun 28, 2021

xiki-tempula deleted the attr branch July 5, 2022 09:55

		>>> dhdl_coul = alchemlyb.concat([extract_dHdl(xvg, T=300) for xvg in
		dataset['Coulomb']])

	'scipy': ('http://docs.scipy.org/doc/scipy/reference/', None)}
	'scipy': ('https://docs.scipy.org/doc/scipy/reference/', None)}

Add attr to the dataframe #129

Add attr to the dataframe #129

Conversation

xiki-tempula commented May 15, 2021 • edited by orbeckst Loading

codecov bot commented May 16, 2021 • edited Loading

Codecov Report

xiki-tempula commented May 21, 2021 • edited Loading

orbeckst left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

xiki-tempula May 28, 2021 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

xiki-tempula Jun 3, 2021 • edited Loading

Choose a reason for hiding this comment

orbeckst commented May 31, 2021

orbeckst left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

xiki-tempula commented Jun 12, 2021

orbeckst commented Jun 14, 2021

orbeckst commented Jun 17, 2021

xiki-tempula commented Jun 17, 2021

orbeckst left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

xiki-tempula commented Jun 19, 2021

orbeckst left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

xiki-tempula Jun 21, 2021 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

xiki-tempula Jun 21, 2021 • edited Loading

Choose a reason for hiding this comment

xiki-tempula commented Jun 21, 2021

orbeckst commented Jun 23, 2021

orbeckst commented Jun 23, 2021

xiki-tempula commented Jun 26, 2021 • edited Loading

orbeckst left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

xiki-tempula commented May 15, 2021 •

edited by orbeckst

Loading

codecov bot commented May 16, 2021 •

edited

Loading

xiki-tempula commented May 21, 2021 •

edited

Loading

xiki-tempula May 28, 2021 •

edited

Loading

xiki-tempula Jun 3, 2021 •

edited

Loading

xiki-tempula Jun 21, 2021 •

edited

Loading

xiki-tempula Jun 21, 2021 •

edited

Loading

xiki-tempula commented Jun 26, 2021 •

edited

Loading