Changes to wrapper to integrate with Bibiserv

[foobarx]

No description provided.

[aweimann]

Why would you replace sort_values with sort_index?

[foobarx]

DataFrame.sort_values() is only available in Pandas 0.17.0 seemingly, and the version in Ubuntu Trusty is 0.13.0. No backports are available either. Could build the Docker image with a later Ubuntu release though. http://pandas.pydata.org/pandas-docs/stable/whatsnew.html#changes-to-sorting-api http://pandas.pydata.org/pandas-docs/version/0.17.0/generated/pandas.DataFrame.sort_values.html https://stackoverflow.com/questions/19332171/difference-between-sort-values-and-sort-index

…

On 2 August 2017 at 13:58, aweimann ***@***.***> wrote: ***@***.**** commented on this pull request. Why would you replace sort_values with sort_index? — You are receiving this because you modified the open/close state. Reply to this email directly, view it on GitHub <#62 (review)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AGRBpGzoYdvoxwEV_Y9QUJaDrxxhzjEDks5sUGSDgaJpZM4OpiGX> .

[aweimann]

Oh I see. The solution is to use bioconda / conda for installing the dependencies. Actually, I'm currently working on making a conda recipe for Traitar itself. Could you use that as well? https://conda.io/docs/install/quick.html

[aweimann]

thanks for the changes; I think I haven't really properly tested this one!

[aweimann]

Why would you do this?

[foobarx]

Bioconda - sounds good. So, we can just change all the deps in the Dockerfile (inc. Traitar).

In traitar/traitar_from_archive.py

<#62 (comment)>: 2nd argument for DF was a DF and not a list

In traitar/hmmer2filtered_best.py

<#62 (comment)>: As discussed - only available in Pandas 0.17.0

…

On 2 August 2017 at 14:14, aweimann ***@***.***> wrote: ***@***.**** commented on this pull request. Oh I see. The solution is to use bioconda / conda for installing the dependencies. Actually, I'm currently working on making a conda recipe for Traitar itself. Could you use that as well? ------------------------------ In traitar/traitar_from_archive.py <#62 (comment)>: > #replace index with cleaned file names sample_cat.index.rename(str, dict([(tf, sfn) for sfn, tf in zip(sample_file_names, namelist)])) - sample_table = pd.DataFrame([sample_file_names, sample_cat.loc[sample_file_names,]]) + sample_table = pd.DataFrame(sample_file_names) + categories = pd.Series(sample_cat.loc[sample_file_names, ]['category'].tolist()) + sample_table['category'] = categories sample_table.columns = ["sample_file_name", "category"] thanks for the changes; I think I haven't really properly tested this one! ------------------------------ In traitar/hmmer2filtered_best.py <#62 (comment)>: > @@ -49,7 +49,7 @@ def aggregate_domain_hits(filtered_df, out_f): #sort by gene identifier and Pfam with open(out_f, 'w') as out_fo: ps.DataFrame(filtered_df.columns).T.to_csv(out_f, sep = "\t", index = False, header = False, mode = 'a') - filtered_df.sort_values(by = ["target name", "query name"], inplace = True) + filtered_df.sort_index(by = ["target name", "query name"], inplace = True) Why would you do this? — You are receiving this because you modified the open/close state. Reply to this email directly, view it on GitHub <#62 (review)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AGRBpKPLLtKx5ytB1QR9qAj_iVosGPSiks5sUGgegaJpZM4OpiGX> .