File cleanup: remove commented print statements and some commented code by zurabksmlp · Pull Request #44 · SMLP-Systems/smlp

zurabksmlp · 2026-02-07T11:08:07Z

This PR contains repository and file cleanup -- some of the unused files got removed and commented-out print statements (and occasionally other commented-out code) has been remoted in files at repo/smlp/src/smlp_py directory.

…mented out code

fbrausse

Nice PR, looks good to me, thanks. Just a few questions:

fbrausse · 2026-02-07T12:30:57Z

src/smlp_py/smlp_precisions.py

+    #true_pred_df = data.frame(true= bin_true_vec, pred=bin_pred_vec)
    #kappa2_res = kappa2(true_pred_df, "unweighted"); 
    #kappa2_res = kappa2(true_pred_df, "squared"); 
-    #cohen_kappa = kappa2_res$value; #pnv(cohen_kappa); 
+    #cohen_kappa = kappa2_res$value 
    #true_pred_categ_df = data.frame(true=factor(bin_true_vec, levels=c(0,1)), 
-    #          pred=factor(bin_pred_vec, levels=c(0,1))); pnv(true_pred_categ_df); 
+    #          pred=factor(bin_pred_vec, levels=c(0,1))) 


Do we still need this commented code?

fbrausse · 2026-02-07T12:31:36Z

src/smlp_py/smlp_precisions.py

+            #pos_expectation = (predicted_negative/samples_count)*true_negative
+            #neg_expectation = (predicted_positive/samples_count)*TP


Same question: do we still need this commented code?

fbrausse · 2026-02-07T12:32:04Z

src/smlp_py/smlp_precisions.py

+            #pos.scores = pos_prob_vec[bin_true_vec_pos]
+            #neg.scores = pos_prob_vec[bin_true_vec_neg]


And again: do we need this?

fbrausse · 2026-02-07T12:32:36Z

src/smlp_py/smlp_query.py

            solver.add(query)
        res = self._modelTermsInst.smlp_solver_check(solver, 'witness_consistency')
-        #res = solver.check(); #print('res', res)
+        #res = solver.check()


Let's remove this commented line as well.

src/smlp_py/smlp_subgroups.py

zurabksmlp · 2026-02-07T13:00:19Z

In most cases, I kept the commented lines because they provide alternative implementation and I was not sure which implementation is best. So these lines are a reminder like the TODO comments that a revision might be relevant. You can see such commented out lines in many places, even functions like # equality @conditional_cache ***@***.*** def smlp_eq(self, term1:smlp.term2, term2:smlp.term2): res1 = op.eq(term1, term2) #res2 = term1 == term2 #assert res1 == res2 return res1 Here you see an alternative implementation and sanity check that is commented out. I am not sure specifically for these functions, but from what I remember using different implementations did affect performance of some of the tests. Yes, in some cases some commented out code can be safely removed such as in the case you have pointed out (that variable ft is never used) but it was not my intention to make a full cleanup. Zurab <https://www.avast.com/sig-email?utm_medium=email&utm_source=link&utm_campaign=sig-email&utm_content=webmail> Virus-free.www.avast.com <https://www.avast.com/sig-email?utm_medium=email&utm_source=link&utm_campaign=sig-email&utm_content=webmail> <#DAB4FAD8-2DD7-40BB-A1B8-4E2AA1F9FDF2>

…

On Sat, Feb 7, 2026 at 2:45 PM Franz Brauße ***@***.***> wrote: ***@***.**** commented on this pull request. Nice PR, looks good to me, thanks. Just a few questions: ------------------------------ In src/smlp_py/smlp_precisions.py <#44 (comment)>: > + #true_pred_df = data.frame(true= bin_true_vec, pred=bin_pred_vec) #kappa2_res = kappa2(true_pred_df, "unweighted"); #kappa2_res = kappa2(true_pred_df, "squared"); - #cohen_kappa = kappa2_res$value; #pnv(cohen_kappa); + #cohen_kappa = kappa2_res$value #true_pred_categ_df = data.frame(true=factor(bin_true_vec, levels=c(0,1)), - # pred=factor(bin_pred_vec, levels=c(0,1))); pnv(true_pred_categ_df); + # pred=factor(bin_pred_vec, levels=c(0,1))) Do we still need this commented code? ------------------------------ In src/smlp_py/smlp_precisions.py <#44 (comment)>: > + #pos_expectation = (predicted_negative/samples_count)*true_negative + #neg_expectation = (predicted_positive/samples_count)*TP Same question: do we still need this commented code? ------------------------------ In src/smlp_py/smlp_precisions.py <#44 (comment)>: > + #pos.scores = pos_prob_vec[bin_true_vec_pos] + #neg.scores = pos_prob_vec[bin_true_vec_neg] And again: do we need this? ------------------------------ In src/smlp_py/smlp_query.py <#44 (comment)>: > if query is not None: solver.add(query) res = self._modelTermsInst.smlp_solver_check(solver, 'witness_consistency') - #res = solver.check(); #print('res', res) + #res = solver.check() Let's remove this commented line as well. ------------------------------ In src/smlp_py/smlp_subgroups.py <#44 (comment)>: > else: - #ft = feat_df[rng_tuple['name']]; #print('ft\n', ft); print(rng_tuple['hi'], rng_tuple['lo']) - #print(rng_tuple); print(rng_tuple$hi); print(rng_tuple$lo) + #ft = feat_df[rng_tuple['name']] I don't think ft is used anywhere, so this commented line should be safe to remove as well. — Reply to this email directly, view it on GitHub <#44 (review)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/BPWPDL7LH3MKPXOCKGLPOVT4KXM45AVCNFSM6AAAAACUKFRUUSVHI2DSMVQWIX3LMV43YUDVNRWFEZLROVSXG5CSMV3GSZLXHMZTONRWHE3DGNBWGE> . You are receiving this because you authored the thread.Message ID: ***@***.***>

zurabksmlp · 2026-02-07T13:04:10Z

By the way , pnv() was a function in EVA that would print the name and the value of a variable (in R language). In Python, I use print('x', x), with pnv() it would be pnv(x). I remember I tried to write such a function in Python since it is very handy, but I could not do it. If you know how to do it, let me know. Zurab <https://www.avast.com/sig-email?utm_medium=email&utm_source=link&utm_campaign=sig-email&utm_content=webmail> Virus-free.www.avast.com <https://www.avast.com/sig-email?utm_medium=email&utm_source=link&utm_campaign=sig-email&utm_content=webmail> <#DAB4FAD8-2DD7-40BB-A1B8-4E2AA1F9FDF2> On Sat, Feb 7, 2026 at 3:00 PM Zurab Khasidashvili < ***@***.***> wrote:

…

In most cases, I kept the commented lines because they provide alternative implementation and I was not sure which implementation is best. So these lines are a reminder like the TODO comments that a revision might be relevant. You can see such commented out lines in many places, even functions like # equality @conditional_cache ***@***.*** def smlp_eq(self, term1:smlp.term2, term2:smlp.term2): res1 = op.eq(term1, term2) #res2 = term1 == term2 #assert res1 == res2 return res1 Here you see an alternative implementation and sanity check that is commented out. I am not sure specifically for these functions, but from what I remember using different implementations did affect performance of some of the tests. Yes, in some cases some commented out code can be safely removed such as in the case you have pointed out (that variable ft is never used) but it was not my intention to make a full cleanup. Zurab <https://www.avast.com/sig-email?utm_medium=email&utm_source=link&utm_campaign=sig-email&utm_content=webmail> Virus-free.www.avast.com <https://www.avast.com/sig-email?utm_medium=email&utm_source=link&utm_campaign=sig-email&utm_content=webmail> <#m_1703351304392304679_DAB4FAD8-2DD7-40BB-A1B8-4E2AA1F9FDF2> On Sat, Feb 7, 2026 at 2:45 PM Franz Brauße ***@***.***> wrote: > ***@***.**** commented on this pull request. > > Nice PR, looks good to me, thanks. Just a few questions: > ------------------------------ > > In src/smlp_py/smlp_precisions.py > <#44 (comment)>: > > > + #true_pred_df = data.frame(true= bin_true_vec, pred=bin_pred_vec) > #kappa2_res = kappa2(true_pred_df, "unweighted"); > #kappa2_res = kappa2(true_pred_df, "squared"); > - #cohen_kappa = kappa2_res$value; #pnv(cohen_kappa); > + #cohen_kappa = kappa2_res$value > #true_pred_categ_df = data.frame(true=factor(bin_true_vec, levels=c(0,1)), > - # pred=factor(bin_pred_vec, levels=c(0,1))); pnv(true_pred_categ_df); > + # pred=factor(bin_pred_vec, levels=c(0,1))) > > Do we still need this commented code? > ------------------------------ > > In src/smlp_py/smlp_precisions.py > <#44 (comment)>: > > > + #pos_expectation = (predicted_negative/samples_count)*true_negative > + #neg_expectation = (predicted_positive/samples_count)*TP > > Same question: do we still need this commented code? > ------------------------------ > > In src/smlp_py/smlp_precisions.py > <#44 (comment)>: > > > + #pos.scores = pos_prob_vec[bin_true_vec_pos] > + #neg.scores = pos_prob_vec[bin_true_vec_neg] > > And again: do we need this? > ------------------------------ > > In src/smlp_py/smlp_query.py > <#44 (comment)>: > > > if query is not None: > solver.add(query) > res = self._modelTermsInst.smlp_solver_check(solver, 'witness_consistency') > - #res = solver.check(); #print('res', res) > + #res = solver.check() > > Let's remove this commented line as well. > ------------------------------ > > In src/smlp_py/smlp_subgroups.py > <#44 (comment)>: > > > else: > - #ft = feat_df[rng_tuple['name']]; #print('ft\n', ft); print(rng_tuple['hi'], rng_tuple['lo']) > - #print(rng_tuple); print(rng_tuple$hi); print(rng_tuple$lo) > + #ft = feat_df[rng_tuple['name']] > > I don't think ft is used anywhere, so this commented line should be safe > to remove as well. > > — > Reply to this email directly, view it on GitHub > <#44 (review)>, > or unsubscribe > <https://github.com/notifications/unsubscribe-auth/BPWPDL7LH3MKPXOCKGLPOVT4KXM45AVCNFSM6AAAAACUKFRUUSVHI2DSMVQWIX3LMV43YUDVNRWFEZLROVSXG5CSMV3GSZLXHMZTONRWHE3DGNBWGE> > . > You are receiving this because you authored the thread.Message ID: > ***@***.***> >

konstantin-korovin · 2026-02-07T13:29:40Z

To print name and value you can use print(f"{var_name = }") https://stackoverflow.com/questions/32000934/print-a-variables-name-and-value/57225950#57225950 foo = 1 bar = 2 print(f"{foo = } {bar =}") produces: foo = 1 bar =2

…

On Sat, 7 Feb 2026 at 13:04, zurabksmlp ***@***.***> wrote: *zurabksmlp* left a comment (SMLP-Systems/smlp#44) <#44 (comment)> By the way , pnv() was a function in EVA that would print the name and the value of a variable (in R language). In Python, I use print('x', x), with pnv() it would be pnv(x). I remember I tried to write such a function in Python since it is very handy, but I could not do it. If you know how to do it, let me know. Zurab < https://www.avast.com/sig-email?utm_medium=email&utm_source=link&utm_campaign=sig-email&utm_content=webmail> Virus-free.www.avast.com < https://www.avast.com/sig-email?utm_medium=email&utm_source=link&utm_campaign=sig-email&utm_content=webmail> <#DAB4FAD8-2DD7-40BB-A1B8-4E2AA1F9FDF2> On Sat, Feb 7, 2026 at 3:00 PM Zurab Khasidashvili < ***@***.***> wrote: > In most cases, I kept the commented lines because they provide alternative > implementation and I was not sure which implementation is best. So these > lines are a reminder like the TODO comments that a revision might be > relevant. You can see such commented out lines in many places, even > functions like > > # equality > @conditional_cache ***@***.*** > def smlp_eq(self, term1:smlp.term2, term2:smlp.term2): > res1 = op.eq(term1, term2) > #res2 = term1 == term2 > #assert res1 == res2 > return res1 > > Here you see an alternative implementation and sanity check that is > commented out. I am not sure specifically for these functions, but from > what I remember using different implementations did affect performance of > some of the tests. > > Yes, in some cases some commented out code can be safely removed such as > in the case you have pointed out (that variable ft is never used) but it > was not my intention to make a full cleanup. > > Zurab > > > < https://www.avast.com/sig-email?utm_medium=email&utm_source=link&utm_campaign=sig-email&utm_content=webmail> > Virus-free.www.avast.com > < https://www.avast.com/sig-email?utm_medium=email&utm_source=link&utm_campaign=sig-email&utm_content=webmail> > <#m_1703351304392304679_DAB4FAD8-2DD7-40BB-A1B8-4E2AA1F9FDF2> > > On Sat, Feb 7, 2026 at 2:45 PM Franz Brauße ***@***.***> > wrote: > >> ***@***.**** commented on this pull request. >> >> Nice PR, looks good to me, thanks. Just a few questions: >> ------------------------------ >> >> In src/smlp_py/smlp_precisions.py >> <#44 (comment)>: >> >> > + #true_pred_df = data.frame(true= bin_true_vec, pred=bin_pred_vec) >> #kappa2_res = kappa2(true_pred_df, "unweighted"); >> #kappa2_res = kappa2(true_pred_df, "squared"); >> - #cohen_kappa = kappa2_res$value; #pnv(cohen_kappa); >> + #cohen_kappa = kappa2_res$value >> #true_pred_categ_df = data.frame(true=factor(bin_true_vec, levels=c(0,1)), >> - # pred=factor(bin_pred_vec, levels=c(0,1))); pnv(true_pred_categ_df); >> + # pred=factor(bin_pred_vec, levels=c(0,1))) >> >> Do we still need this commented code? >> ------------------------------ >> >> In src/smlp_py/smlp_precisions.py >> <#44 (comment)>: >> >> > + #pos_expectation = (predicted_negative/samples_count)*true_negative >> + #neg_expectation = (predicted_positive/samples_count)*TP >> >> Same question: do we still need this commented code? >> ------------------------------ >> >> In src/smlp_py/smlp_precisions.py >> <#44 (comment)>: >> >> > + #pos.scores = pos_prob_vec[bin_true_vec_pos] >> + #neg.scores = pos_prob_vec[bin_true_vec_neg] >> >> And again: do we need this? >> ------------------------------ >> >> In src/smlp_py/smlp_query.py >> <#44 (comment)>: >> >> > if query is not None: >> solver.add(query) >> res = self._modelTermsInst.smlp_solver_check(solver, 'witness_consistency') >> - #res = solver.check(); #print('res', res) >> + #res = solver.check() >> >> Let's remove this commented line as well. >> ------------------------------ >> >> In src/smlp_py/smlp_subgroups.py >> <#44 (comment)>: >> >> > else: >> - #ft = feat_df[rng_tuple['name']]; #print('ft\n', ft); print(rng_tuple['hi'], rng_tuple['lo']) >> - #print(rng_tuple); print(rng_tuple$hi); print(rng_tuple$lo) >> + #ft = feat_df[rng_tuple['name']] >> >> I don't think ft is used anywhere, so this commented line should be safe >> to remove as well. >> >> — >> Reply to this email directly, view it on GitHub >> < #44 (review)>, >> or unsubscribe >> < https://github.com/notifications/unsubscribe-auth/BPWPDL7LH3MKPXOCKGLPOVT4KXM45AVCNFSM6AAAAACUKFRUUSVHI2DSMVQWIX3LMV43YUDVNRWFEZLROVSXG5CSMV3GSZLXHMZTONRWHE3DGNBWGE> >> . >> You are receiving this because you authored the thread.Message ID: >> ***@***.***> >> > — Reply to this email directly, view it on GitHub <#44 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AIVYHTJ4WB7BKPQCXPT3QIL4KXPF7AVCNFSM6AAAAACUKFRUUSVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZTQNRUGQ4DKMJWGA> . You are receiving this because you are subscribed to this thread.Message ID: ***@***.***>

fbrausse · 2026-02-07T13:42:16Z

@zurabksmlp Thanks. I did merge your previous PR #43, so now Github lets us know that there are conflicts between master and this PR #44. Could I ask you to please resolve those conflicts? You could either use the web interface or try to do that locally via git rebase master file-cleanup2.

zurabksmlp · 2026-02-07T13:46:15Z

Thanks Konstantin, it is a good idea but still requires typing extra characters like f" {} " and is not significantly quicker. but it is a good idea because maybe one can use *kwargs for that purpose. If I simply define: def pnv(x): print(f"x = {x}") y = 7 pnv(y) then x = 7 is printed instead of y = 7. <https://www.avast.com/sig-email?utm_medium=email&utm_source=link&utm_campaign=sig-email&utm_content=webmail> Virus-free.www.avast.com <https://www.avast.com/sig-email?utm_medium=email&utm_source=link&utm_campaign=sig-email&utm_content=webmail> <#DAB4FAD8-2DD7-40BB-A1B8-4E2AA1F9FDF2> On Sat, Feb 7, 2026 at 3:30 PM konstantin-korovin ***@***.***> wrote:

…

*konstantin-korovin* left a comment (SMLP-Systems/smlp#44) <#44 (comment)> To print name and value you can use print(f"{var_name = }") https://stackoverflow.com/questions/32000934/print-a-variables-name-and-value/57225950#57225950 foo = 1 bar = 2 print(f"{foo = } {bar =}") produces: foo = 1 bar =2 On Sat, 7 Feb 2026 at 13:04, zurabksmlp ***@***.***> wrote: > *zurabksmlp* left a comment (SMLP-Systems/smlp#44) > <#44 (comment)> > By the way , pnv() was a function in EVA that would print the name and the > value of a variable (in R language). > In Python, I use print('x', x), with pnv() it would be pnv(x). > > I remember I tried to write such a function in Python since it is very > handy, but I could not do it. > If you know how to do it, let me know. > > Zurab > > < > https://www.avast.com/sig-email?utm_medium=email&utm_source=link&utm_campaign=sig-email&utm_content=webmail> > > Virus-free.www.avast.com > < > https://www.avast.com/sig-email?utm_medium=email&utm_source=link&utm_campaign=sig-email&utm_content=webmail> > > <#DAB4FAD8-2DD7-40BB-A1B8-4E2AA1F9FDF2> > > On Sat, Feb 7, 2026 at 3:00 PM Zurab Khasidashvili < > ***@***.***> wrote: > > > In most cases, I kept the commented lines because they provide > alternative > > implementation and I was not sure which implementation is best. So these > > lines are a reminder like the TODO comments that a revision might be > > relevant. You can see such commented out lines in many places, even > > functions like > > > > # equality > > @conditional_cache ***@***.*** > > def smlp_eq(self, term1:smlp.term2, term2:smlp.term2): > > res1 = op.eq(term1, term2) > > #res2 = term1 == term2 > > #assert res1 == res2 > > return res1 > > > > Here you see an alternative implementation and sanity check that is > > commented out. I am not sure specifically for these functions, but from > > what I remember using different implementations did affect performance > of > > some of the tests. > > > > Yes, in some cases some commented out code can be safely removed such as > > in the case you have pointed out (that variable ft is never used) but it > > was not my intention to make a full cleanup. > > > > Zurab > > > > > > < > https://www.avast.com/sig-email?utm_medium=email&utm_source=link&utm_campaign=sig-email&utm_content=webmail> > > > Virus-free.www.avast.com > > < > https://www.avast.com/sig-email?utm_medium=email&utm_source=link&utm_campaign=sig-email&utm_content=webmail> > > > <#m_1703351304392304679_DAB4FAD8-2DD7-40BB-A1B8-4E2AA1F9FDF2> > > > > On Sat, Feb 7, 2026 at 2:45 PM Franz Brauße ***@***.***> > > wrote: > > > >> ***@***.**** commented on this pull request. > >> > >> Nice PR, looks good to me, thanks. Just a few questions: > >> ------------------------------ > >> > >> In src/smlp_py/smlp_precisions.py > >> <#44 (comment)>: > >> > >> > + #true_pred_df = data.frame(true= bin_true_vec, pred=bin_pred_vec) > >> #kappa2_res = kappa2(true_pred_df, "unweighted"); > >> #kappa2_res = kappa2(true_pred_df, "squared"); > >> - #cohen_kappa = kappa2_res$value; #pnv(cohen_kappa); > >> + #cohen_kappa = kappa2_res$value > >> #true_pred_categ_df = data.frame(true=factor(bin_true_vec, > levels=c(0,1)), > >> - # pred=factor(bin_pred_vec, levels=c(0,1))); pnv(true_pred_categ_df); > >> + # pred=factor(bin_pred_vec, levels=c(0,1))) > >> > >> Do we still need this commented code? > >> ------------------------------ > >> > >> In src/smlp_py/smlp_precisions.py > >> <#44 (comment)>: > >> > >> > + #pos_expectation = (predicted_negative/samples_count)*true_negative > >> + #neg_expectation = (predicted_positive/samples_count)*TP > >> > >> Same question: do we still need this commented code? > >> ------------------------------ > >> > >> In src/smlp_py/smlp_precisions.py > >> <#44 (comment)>: > >> > >> > + #pos.scores = pos_prob_vec[bin_true_vec_pos] > >> + #neg.scores = pos_prob_vec[bin_true_vec_neg] > >> > >> And again: do we need this? > >> ------------------------------ > >> > >> In src/smlp_py/smlp_query.py > >> <#44 (comment)>: > >> > >> > if query is not None: > >> solver.add(query) > >> res = self._modelTermsInst.smlp_solver_check(solver, > 'witness_consistency') > >> - #res = solver.check(); #print('res', res) > >> + #res = solver.check() > >> > >> Let's remove this commented line as well. > >> ------------------------------ > >> > >> In src/smlp_py/smlp_subgroups.py > >> <#44 (comment)>: > >> > >> > else: > >> - #ft = feat_df[rng_tuple['name']]; #print('ft\n', ft); > print(rng_tuple['hi'], rng_tuple['lo']) > >> - #print(rng_tuple); print(rng_tuple$hi); print(rng_tuple$lo) > >> + #ft = feat_df[rng_tuple['name']] > >> > >> I don't think ft is used anywhere, so this commented line should be > safe > >> to remove as well. > >> > >> — > >> Reply to this email directly, view it on GitHub > >> < > #44 (review)>, > > >> or unsubscribe > >> < > https://github.com/notifications/unsubscribe-auth/BPWPDL7LH3MKPXOCKGLPOVT4KXM45AVCNFSM6AAAAACUKFRUUSVHI2DSMVQWIX3LMV43YUDVNRWFEZLROVSXG5CSMV3GSZLXHMZTONRWHE3DGNBWGE> > > >> . > >> You are receiving this because you authored the thread.Message ID: > >> ***@***.***> > >> > > > > — > Reply to this email directly, view it on GitHub > <#44 (comment)>, > or unsubscribe > < https://github.com/notifications/unsubscribe-auth/AIVYHTJ4WB7BKPQCXPT3QIL4KXPF7AVCNFSM6AAAAACUKFRUUSVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZTQNRUGQ4DKMJWGA> > . > You are receiving this because you are subscribed to this thread.Message > ID: ***@***.***> > — Reply to this email directly, view it on GitHub <#44 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/BPWPDL7EVLQM5OFHLMXFXTL4KXSFXAVCNFSM6AAAAACUKFRUUSVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZTQNRUGUYTEMJSGM> . You are receiving this because you authored the thread.Message ID: ***@***.***>

zurabksmlp · 2026-02-07T13:59:22Z

Done -- executed: (venv_3.12.3) ***@***.***:~/smlp/repo/smlp/src$ git rebase master file-cleanup2 Current branch file-cleanup2 is up to date. Zurab <https://www.avast.com/sig-email?utm_medium=email&utm_source=link&utm_campaign=sig-email&utm_content=webmail> Virus-free.www.avast.com <https://www.avast.com/sig-email?utm_medium=email&utm_source=link&utm_campaign=sig-email&utm_content=webmail> <#DAB4FAD8-2DD7-40BB-A1B8-4E2AA1F9FDF2> On Sat, Feb 7, 2026 at 3:46 PM Zurab Khasidashvili < ***@***.***> wrote:

…

Thanks Konstantin, it is a good idea but still requires typing extra characters like f" {} " and is not significantly quicker. but it is a good idea because maybe one can use *kwargs for that purpose. If I simply define: def pnv(x): print(f"x = {x}") y = 7 pnv(y) then x = 7 is printed instead of y = 7. <https://www.avast.com/sig-email?utm_medium=email&utm_source=link&utm_campaign=sig-email&utm_content=webmail> Virus-free.www.avast.com <https://www.avast.com/sig-email?utm_medium=email&utm_source=link&utm_campaign=sig-email&utm_content=webmail> <#m_-8140363696942388275_DAB4FAD8-2DD7-40BB-A1B8-4E2AA1F9FDF2> On Sat, Feb 7, 2026 at 3:30 PM konstantin-korovin < ***@***.***> wrote: > *konstantin-korovin* left a comment (SMLP-Systems/smlp#44) > <#44 (comment)> > To print name and value you can use > > print(f"{var_name = }") > > > > https://stackoverflow.com/questions/32000934/print-a-variables-name-and-value/57225950#57225950 > > > foo = 1 > bar = 2 > > print(f"{foo = } {bar =}") > > produces: > > foo = 1 bar =2 > > > On Sat, 7 Feb 2026 at 13:04, zurabksmlp ***@***.***> wrote: > > > *zurabksmlp* left a comment (SMLP-Systems/smlp#44) > > <#44 (comment)> > > By the way , pnv() was a function in EVA that would print the name and > the > > value of a variable (in R language). > > In Python, I use print('x', x), with pnv() it would be pnv(x). > > > > I remember I tried to write such a function in Python since it is very > > handy, but I could not do it. > > If you know how to do it, let me know. > > > > Zurab > > > > < > > > https://www.avast.com/sig-email?utm_medium=email&utm_source=link&utm_campaign=sig-email&utm_content=webmail> > > > > > Virus-free.www.avast.com > > < > > > https://www.avast.com/sig-email?utm_medium=email&utm_source=link&utm_campaign=sig-email&utm_content=webmail> > > > > > <#DAB4FAD8-2DD7-40BB-A1B8-4E2AA1F9FDF2> > > > > On Sat, Feb 7, 2026 at 3:00 PM Zurab Khasidashvili < > > ***@***.***> wrote: > > > > > In most cases, I kept the commented lines because they provide > > alternative > > > implementation and I was not sure which implementation is best. So > these > > > lines are a reminder like the TODO comments that a revision might be > > > relevant. You can see such commented out lines in many places, even > > > functions like > > > > > > # equality > > > @conditional_cache ***@***.*** > > > def smlp_eq(self, term1:smlp.term2, term2:smlp.term2): > > > res1 = op.eq(term1, term2) > > > #res2 = term1 == term2 > > > #assert res1 == res2 > > > return res1 > > > > > > Here you see an alternative implementation and sanity check that is > > > commented out. I am not sure specifically for these functions, but > from > > > what I remember using different implementations did affect > performance > > of > > > some of the tests. > > > > > > Yes, in some cases some commented out code can be safely removed such > as > > > in the case you have pointed out (that variable ft is never used) but > it > > > was not my intention to make a full cleanup. > > > > > > Zurab > > > > > > > > > < > > > https://www.avast.com/sig-email?utm_medium=email&utm_source=link&utm_campaign=sig-email&utm_content=webmail> > > > > > > Virus-free.www.avast.com > > > < > > > https://www.avast.com/sig-email?utm_medium=email&utm_source=link&utm_campaign=sig-email&utm_content=webmail> > > > > > > <#m_1703351304392304679_DAB4FAD8-2DD7-40BB-A1B8-4E2AA1F9FDF2> > > > > > > On Sat, Feb 7, 2026 at 2:45 PM Franz Brauße ***@***.***> > > > wrote: > > > > > >> ***@***.**** commented on this pull request. > > >> > > >> Nice PR, looks good to me, thanks. Just a few questions: > > >> ------------------------------ > > >> > > >> In src/smlp_py/smlp_precisions.py > > >> <#44 (comment)>: > > > >> > > >> > + #true_pred_df = data.frame(true= bin_true_vec, > pred=bin_pred_vec) > > >> #kappa2_res = kappa2(true_pred_df, "unweighted"); > > >> #kappa2_res = kappa2(true_pred_df, "squared"); > > >> - #cohen_kappa = kappa2_res$value; #pnv(cohen_kappa); > > >> + #cohen_kappa = kappa2_res$value > > >> #true_pred_categ_df = data.frame(true=factor(bin_true_vec, > > levels=c(0,1)), > > >> - # pred=factor(bin_pred_vec, levels=c(0,1))); > pnv(true_pred_categ_df); > > >> + # pred=factor(bin_pred_vec, levels=c(0,1))) > > >> > > >> Do we still need this commented code? > > >> ------------------------------ > > >> > > >> In src/smlp_py/smlp_precisions.py > > >> <#44 (comment)>: > > > >> > > >> > + #pos_expectation = > (predicted_negative/samples_count)*true_negative > > >> + #neg_expectation = (predicted_positive/samples_count)*TP > > >> > > >> Same question: do we still need this commented code? > > >> ------------------------------ > > >> > > >> In src/smlp_py/smlp_precisions.py > > >> <#44 (comment)>: > > > >> > > >> > + #pos.scores = pos_prob_vec[bin_true_vec_pos] > > >> + #neg.scores = pos_prob_vec[bin_true_vec_neg] > > >> > > >> And again: do we need this? > > >> ------------------------------ > > >> > > >> In src/smlp_py/smlp_query.py > > >> <#44 (comment)>: > > > >> > > >> > if query is not None: > > >> solver.add(query) > > >> res = self._modelTermsInst.smlp_solver_check(solver, > > 'witness_consistency') > > >> - #res = solver.check(); #print('res', res) > > >> + #res = solver.check() > > >> > > >> Let's remove this commented line as well. > > >> ------------------------------ > > >> > > >> In src/smlp_py/smlp_subgroups.py > > >> <#44 (comment)>: > > > >> > > >> > else: > > >> - #ft = feat_df[rng_tuple['name']]; #print('ft\n', ft); > > print(rng_tuple['hi'], rng_tuple['lo']) > > >> - #print(rng_tuple); print(rng_tuple$hi); print(rng_tuple$lo) > > >> + #ft = feat_df[rng_tuple['name']] > > >> > > >> I don't think ft is used anywhere, so this commented line should be > > safe > > >> to remove as well. > > >> > > >> — > > >> Reply to this email directly, view it on GitHub > > >> < > > > #44 (review)>, > > > > > >> or unsubscribe > > >> < > > > https://github.com/notifications/unsubscribe-auth/BPWPDL7LH3MKPXOCKGLPOVT4KXM45AVCNFSM6AAAAACUKFRUUSVHI2DSMVQWIX3LMV43YUDVNRWFEZLROVSXG5CSMV3GSZLXHMZTONRWHE3DGNBWGE> > > > > > >> . > > >> You are receiving this because you authored the thread.Message ID: > > >> ***@***.***> > > >> > > > > > > > — > > Reply to this email directly, view it on GitHub > > <#44 (comment)>, > > > or unsubscribe > > < > https://github.com/notifications/unsubscribe-auth/AIVYHTJ4WB7BKPQCXPT3QIL4KXPF7AVCNFSM6AAAAACUKFRUUSVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZTQNRUGQ4DKMJWGA> > > > . > > You are receiving this because you are subscribed to this > thread.Message > > ID: ***@***.***> > > > > — > Reply to this email directly, view it on GitHub > <#44 (comment)>, > or unsubscribe > <https://github.com/notifications/unsubscribe-auth/BPWPDL7EVLQM5OFHLMXFXTL4KXSFXAVCNFSM6AAAAACUKFRUUSVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZTQNRUGUYTEMJSGM> > . > You are receiving this because you authored the thread.Message ID: > ***@***.***> >

konstantin-korovin · 2026-02-07T14:11:57Z

Variable names are not passed to functions so you can not access "y" in def pnv(x): when call pnv(y). Not sure how to simplify print(f"{x = }") further. It is still better than print('x', x) as it is less error prone.

…

On Sat, 7 Feb 2026 at 13:46, zurabksmlp ***@***.***> wrote: *zurabksmlp* left a comment (SMLP-Systems/smlp#44) <#44 (comment)> Thanks Konstantin, it is a good idea but still requires typing extra characters like f" {} " and is not significantly quicker. but it is a good idea because maybe one can use *kwargs for that purpose. If I simply define: def pnv(x): print(f"x = {x}") y = 7 pnv(y) then x = 7 is printed instead of y = 7. < https://www.avast.com/sig-email?utm_medium=email&utm_source=link&utm_campaign=sig-email&utm_content=webmail> Virus-free.www.avast.com < https://www.avast.com/sig-email?utm_medium=email&utm_source=link&utm_campaign=sig-email&utm_content=webmail> <#DAB4FAD8-2DD7-40BB-A1B8-4E2AA1F9FDF2> On Sat, Feb 7, 2026 at 3:30 PM konstantin-korovin ***@***.***> wrote: > *konstantin-korovin* left a comment (SMLP-Systems/smlp#44) > <#44 (comment)> > To print name and value you can use > > print(f"{var_name = }") > > > > https://stackoverflow.com/questions/32000934/print-a-variables-name-and-value/57225950#57225950 > > > foo = 1 > bar = 2 > > print(f"{foo = } {bar =}") > > produces: > > foo = 1 bar =2 > > > On Sat, 7 Feb 2026 at 13:04, zurabksmlp ***@***.***> wrote: > > > *zurabksmlp* left a comment (SMLP-Systems/smlp#44) > > <#44 (comment)> > > By the way , pnv() was a function in EVA that would print the name and > the > > value of a variable (in R language). > > In Python, I use print('x', x), with pnv() it would be pnv(x). > > > > I remember I tried to write such a function in Python since it is very > > handy, but I could not do it. > > If you know how to do it, let me know. > > > > Zurab > > > > < > > > https://www.avast.com/sig-email?utm_medium=email&utm_source=link&utm_campaign=sig-email&utm_content=webmail> > > > > > Virus-free.www.avast.com > > < > > > https://www.avast.com/sig-email?utm_medium=email&utm_source=link&utm_campaign=sig-email&utm_content=webmail> > > > > > <#DAB4FAD8-2DD7-40BB-A1B8-4E2AA1F9FDF2> > > > > On Sat, Feb 7, 2026 at 3:00 PM Zurab Khasidashvili < > > ***@***.***> wrote: > > > > > In most cases, I kept the commented lines because they provide > > alternative > > > implementation and I was not sure which implementation is best. So > these > > > lines are a reminder like the TODO comments that a revision might be > > > relevant. You can see such commented out lines in many places, even > > > functions like > > > > > > # equality > > > @conditional_cache ***@***.*** > > > def smlp_eq(self, term1:smlp.term2, term2:smlp.term2): > > > res1 = op.eq(term1, term2) > > > #res2 = term1 == term2 > > > #assert res1 == res2 > > > return res1 > > > > > > Here you see an alternative implementation and sanity check that is > > > commented out. I am not sure specifically for these functions, but > from > > > what I remember using different implementations did affect performance > > of > > > some of the tests. > > > > > > Yes, in some cases some commented out code can be safely removed such > as > > > in the case you have pointed out (that variable ft is never used) but > it > > > was not my intention to make a full cleanup. > > > > > > Zurab > > > > > > > > > < > > > https://www.avast.com/sig-email?utm_medium=email&utm_source=link&utm_campaign=sig-email&utm_content=webmail> > > > > > > Virus-free.www.avast.com > > > < > > > https://www.avast.com/sig-email?utm_medium=email&utm_source=link&utm_campaign=sig-email&utm_content=webmail> > > > > > > <#m_1703351304392304679_DAB4FAD8-2DD7-40BB-A1B8-4E2AA1F9FDF2> > > > > > > On Sat, Feb 7, 2026 at 2:45 PM Franz Brauße ***@***.***> > > > wrote: > > > > > >> ***@***.**** commented on this pull request. > > >> > > >> Nice PR, looks good to me, thanks. Just a few questions: > > >> ------------------------------ > > >> > > >> In src/smlp_py/smlp_precisions.py > > >> < #44 (comment)>: > > > >> > > >> > + #true_pred_df = data.frame(true= bin_true_vec, pred=bin_pred_vec) > > >> #kappa2_res = kappa2(true_pred_df, "unweighted"); > > >> #kappa2_res = kappa2(true_pred_df, "squared"); > > >> - #cohen_kappa = kappa2_res$value; #pnv(cohen_kappa); > > >> + #cohen_kappa = kappa2_res$value > > >> #true_pred_categ_df = data.frame(true=factor(bin_true_vec, > > levels=c(0,1)), > > >> - # pred=factor(bin_pred_vec, levels=c(0,1))); > pnv(true_pred_categ_df); > > >> + # pred=factor(bin_pred_vec, levels=c(0,1))) > > >> > > >> Do we still need this commented code? > > >> ------------------------------ > > >> > > >> In src/smlp_py/smlp_precisions.py > > >> < #44 (comment)>: > > > >> > > >> > + #pos_expectation = > (predicted_negative/samples_count)*true_negative > > >> + #neg_expectation = (predicted_positive/samples_count)*TP > > >> > > >> Same question: do we still need this commented code? > > >> ------------------------------ > > >> > > >> In src/smlp_py/smlp_precisions.py > > >> < #44 (comment)>: > > > >> > > >> > + #pos.scores = pos_prob_vec[bin_true_vec_pos] > > >> + #neg.scores = pos_prob_vec[bin_true_vec_neg] > > >> > > >> And again: do we need this? > > >> ------------------------------ > > >> > > >> In src/smlp_py/smlp_query.py > > >> < #44 (comment)>: > > > >> > > >> > if query is not None: > > >> solver.add(query) > > >> res = self._modelTermsInst.smlp_solver_check(solver, > > 'witness_consistency') > > >> - #res = solver.check(); #print('res', res) > > >> + #res = solver.check() > > >> > > >> Let's remove this commented line as well. > > >> ------------------------------ > > >> > > >> In src/smlp_py/smlp_subgroups.py > > >> < #44 (comment)>: > > > >> > > >> > else: > > >> - #ft = feat_df[rng_tuple['name']]; #print('ft\n', ft); > > print(rng_tuple['hi'], rng_tuple['lo']) > > >> - #print(rng_tuple); print(rng_tuple$hi); print(rng_tuple$lo) > > >> + #ft = feat_df[rng_tuple['name']] > > >> > > >> I don't think ft is used anywhere, so this commented line should be > > safe > > >> to remove as well. > > >> > > >> — > > >> Reply to this email directly, view it on GitHub > > >> < > > > #44 (review)>, > > > > > >> or unsubscribe > > >> < > > > https://github.com/notifications/unsubscribe-auth/BPWPDL7LH3MKPXOCKGLPOVT4KXM45AVCNFSM6AAAAACUKFRUUSVHI2DSMVQWIX3LMV43YUDVNRWFEZLROVSXG5CSMV3GSZLXHMZTONRWHE3DGNBWGE> > > > > > >> . > > >> You are receiving this because you authored the thread.Message ID: > > >> ***@***.***> > > >> > > > > > > > — > > Reply to this email directly, view it on GitHub > > <#44 (comment)>, > > or unsubscribe > > < > https://github.com/notifications/unsubscribe-auth/AIVYHTJ4WB7BKPQCXPT3QIL4KXPF7AVCNFSM6AAAAACUKFRUUSVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZTQNRUGQ4DKMJWGA> > > > . > > You are receiving this because you are subscribed to this thread.Message > > ID: ***@***.***> > > > > — > Reply to this email directly, view it on GitHub > <#44 (comment)>, > or unsubscribe > < https://github.com/notifications/unsubscribe-auth/BPWPDL7EVLQM5OFHLMXFXTL4KXSFXAVCNFSM6AAAAACUKFRUUSVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZTQNRUGUYTEMJSGM> > . > You are receiving this because you authored the thread.Message ID: > ***@***.***> > — Reply to this email directly, view it on GitHub <#44 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AIVYHTMR3VIXXBDG5CPGLCD4KXUD3AVCNFSM6AAAAACUKFRUUSVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZTQNRUGUZTCNZTGI> . You are receiving this because you commented.Message ID: ***@***.***>

fbrausse · 2026-02-07T14:15:57Z

Done

OK, so there were no conflicts when rebasing, good. Please do git push -f to overwrite the remote branch (which this PR tracks) with the contents of your local branch.

mdmitry1 · 2026-02-07T14:22:07Z

Regression passed.
Open issues:

Merge conflicts in src/smlp_py/smlp_optimize.py and src/smlp_py/smlp_terms.py
Need to decide if we'd like to move from tkdiff to kdiff3 and if yes, should tkdiff->kdiff3 change be part of this pull request

fbrausse · 2026-02-07T14:25:31Z

2. tkdiff->kdiff3 change

I'm fine with that change, don't care about the graphical tool default settings. Eventually, those should go away and get replaced with the system standard, e.g., xdg-open or similar.

fbrausse · 2026-02-07T14:28:33Z

replaced with the system standard

Or, alternatively, these settings could become part of a ~/.smlprc or similar local config file. No need to have them hardcoded.

zurabksmlp · 2026-02-07T14:56:41Z

I have run: (venv_3.12.3) ***@***.***:~/smlp/repo/smlp/src$ git push -f Everything up-to-date <https://www.avast.com/sig-email?utm_medium=email&utm_source=link&utm_campaign=sig-email&utm_content=webmail> Virus-free.www.avast.com <https://www.avast.com/sig-email?utm_medium=email&utm_source=link&utm_campaign=sig-email&utm_content=webmail> <#DAB4FAD8-2DD7-40BB-A1B8-4E2AA1F9FDF2>

…

On Sat, Feb 7, 2026 at 4:28 PM Franz Brauße ***@***.***> wrote: *fbrausse* left a comment (SMLP-Systems/smlp#44) <#44 (comment)> replaced with the system standard Or, alternatively, these settings could become part of a ~/.smlprc or similar local config file. No need to have them hardcoded. — Reply to this email directly, view it on GitHub <#44 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/BPWPDLYQVTXMVAVIY76BAPT4KXZCPAVCNFSM6AAAAACUKFRUUSVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZTQNRUGYYDINRWGM> . You are receiving this because you were mentioned.Message ID: ***@***.***>

fbrausse · 2026-02-07T15:35:18Z

@zurabksmlp No, that didn't work, as you can easily check on the Website: #44

Please run the same sequence of commands with an up-to-date master branch.

fbrausse

Merging master into this branch left two commented debug prints. Due to this merge, I recommend re-running the regression suite.

src/smlp_py/smlp_optimize.py

zurabksmlp · 2026-02-08T13:25:43Z

I do not mind removing these two print statements, but is it worth it? Because I will need to again create a PR for file-cleanup2? And Dmitry will need to approve again? <https://www.avast.com/sig-email?utm_medium=email&utm_source=link&utm_campaign=sig-email&utm_content=webmail> Virus-free.www.avast.com <https://www.avast.com/sig-email?utm_medium=email&utm_source=link&utm_campaign=sig-email&utm_content=webmail> <#DAB4FAD8-2DD7-40BB-A1B8-4E2AA1F9FDF2>

…

On Sun, Feb 8, 2026 at 12:39 PM Franz Brauße ***@***.***> wrote: ***@***.**** commented on this pull request. Merging master into this branch left two commented debug prints. Due to this merge, I recommend re-running the regression suite. ------------------------------ In src/smlp_py/smlp_optimize.py <#44 (comment)>: > @@ -393,8 +393,13 @@ def optimize_single_objective_eager(self, model_full_term_dict:dict, objv_name:s # compute and return max value of l_stab within P assert len(P) > 0 - l_res = max([tup[3] for tup in P_eager]) - assert l_res == P[-1]['threshold_lo_scaled'] + #print('P', P) Let's remove this commented print statement as well. ------------------------------ In src/smlp_py/smlp_optimize.py <#44 (comment)>: > @@ -393,8 +393,13 @@ def optimize_single_objective_eager(self, model_full_term_dict:dict, objv_name:s # compute and return max value of l_stab within P assert len(P) > 0 - l_res = max([tup[3] for tup in P_eager]) - assert l_res == P[-1]['threshold_lo_scaled'] + #print('P', P) + l_res = max([tup[3] for tup in P_eager]); #print('l_res', l_res, 'threshold_lo', P[-1]['threshold_lo']) + if scale_objectives or objv_bounds is None: + assert l_res == P[-1]['threshold_lo_scaled'] + else: + assert l_res == P[-1]['threshold_lo'] + #print('optimize_single_objective_eager and with l_res', l_res, flush=True) And this one, please. — Reply to this email directly, view it on GitHub <#44 (review)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/BPWPDLZYMZZOWC3UR5HBNHT4K4G7VAVCNFSM6AAAAACUKFRUUSVHI2DSMVQWIX3LMV43YUDVNRWFEZLROVSXG5CSMV3GSZLXHMZTONRZGYYTSMZWGY> . You are receiving this because you were mentioned.Message ID: ***@***.***>

fbrausse · 2026-02-08T17:49:52Z

I do not mind removing these two print statements, but is it worth it?
Because I will need to again create a PR for file-cleanup2? And Dmitry
will need to approve again?

Yes, it is worth removing those two lines in this PR. Noone has approved it, yet, and running the regression suite should be done again anyway since the master merge did result in semantic changes, no?

You can just add a commit to the existing file-cleanup2 branch and it will be reflected in this PR. Just switch over to that branch, make sure it is in sync with the remote one on Github, add a commit that removes those two lines and push it back.

…t line in smlp_subgroups.py

zurabksmlp · 2026-02-08T19:05:45Z

OK, done now. Three prints removed from smlp_optimize.py and one commented out line in smlp_subgroups.py (based on your earlier feedback). <https://www.avast.com/sig-email?utm_medium=email&utm_source=link&utm_campaign=sig-email&utm_content=webmail> Virus-free.www.avast.com <https://www.avast.com/sig-email?utm_medium=email&utm_source=link&utm_campaign=sig-email&utm_content=webmail> <#DAB4FAD8-2DD7-40BB-A1B8-4E2AA1F9FDF2>

…

On Sun, Feb 8, 2026 at 7:50 PM Franz Brauße ***@***.***> wrote: *fbrausse* left a comment (SMLP-Systems/smlp#44) <#44 (comment)> I do not mind removing these two print statements, but is it worth it? Because I will need to again create a PR for file-cleanup2? And Dmitry will need to approve again? Yes, it is worth removing those two lines *in this PR*. Noone has approved it, yet, and running the regression suite should be done again anyway since the master merge did result in semantic changes, no? You can just add a commit to the existing file-cleanup2 branch and it will be reflected in this PR. Just switch over to that branch, make sure it is in sync with the remote one on Github, add a commit that removes those two lines and push it back. — Reply to this email directly, view it on GitHub <#44 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/BPWPDL2EHIMKHQWOGCOLJNL4K5ZNNAVCNFSM6AAAAACUKFRUUSVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZTQNRXG4ZTQOBZGQ> . You are receiving this because you were mentioned.Message ID: ***@***.***>

fbrausse · 2026-02-08T19:50:12Z

Many thanks, @zurabksmlp! I'll run regressions locally again.

zurabksmlp added 2 commits February 6, 2026 12:17

removing printings from all files. in rare occasions also removed com…

7bdd7fa

…mented out code

renoving print statements

68b7cd2

fbrausse reviewed Feb 7, 2026

View reviewed changes

fbrausse changed the title ~~File cleanup2 -- updated~~ File cleanup: remove commented print statements and some commented code Feb 7, 2026

Merge branch 'master' into file-cleanup2

a2f0e1c

fbrausse reviewed Feb 8, 2026

View reviewed changes

src/smlp_py/smlp_optimize.py Outdated Show resolved Hide resolved

src/smlp_py/smlp_optimize.py Outdated Show resolved Hide resolved

removed a few print statements in smlp_optimize.py and a commented-ou…

a187c35

…t line in smlp_subgroups.py

mdmitry1 mentioned this pull request Feb 8, 2026

SMLP python3.13 version: includes master update and docker support #48

Draft

		#pos_expectation = (predicted_negative/samples_count)*true_negative
		#neg_expectation = (predicted_positive/samples_count)*TP

		#pos.scores = pos_prob_vec[bin_true_vec_pos]
		#neg.scores = pos_prob_vec[bin_true_vec_neg]

Conversation

zurabksmlp commented Feb 7, 2026

Uh oh!

fbrausse left a comment

Choose a reason for hiding this comment

Uh oh!

fbrausse Feb 7, 2026

Choose a reason for hiding this comment

Uh oh!

fbrausse Feb 7, 2026

Choose a reason for hiding this comment

Uh oh!

fbrausse Feb 7, 2026

Choose a reason for hiding this comment

Uh oh!

fbrausse Feb 7, 2026

Choose a reason for hiding this comment

Uh oh!

Uh oh!

zurabksmlp commented Feb 7, 2026 via email

Uh oh!

zurabksmlp commented Feb 7, 2026 via email

Uh oh!

konstantin-korovin commented Feb 7, 2026 via email

Uh oh!

fbrausse commented Feb 7, 2026

Uh oh!

zurabksmlp commented Feb 7, 2026 via email

Uh oh!

zurabksmlp commented Feb 7, 2026 via email

Uh oh!

konstantin-korovin commented Feb 7, 2026 via email

Uh oh!

fbrausse commented Feb 7, 2026

Uh oh!

mdmitry1 commented Feb 7, 2026

Uh oh!

fbrausse commented Feb 7, 2026

Uh oh!

fbrausse commented Feb 7, 2026

Uh oh!

zurabksmlp commented Feb 7, 2026 via email

Uh oh!

fbrausse commented Feb 7, 2026

Uh oh!

fbrausse left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

zurabksmlp commented Feb 8, 2026 via email

Uh oh!

fbrausse commented Feb 8, 2026

Uh oh!

zurabksmlp commented Feb 8, 2026 via email

Uh oh!

fbrausse commented Feb 8, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants