ENH Normalize output format

This makes the result more reproducible (and should fix test failures)
BigDataBiology · Apr 25, 2024 · 4acde50 · 4acde50
1 parent 0c1bf81
commit 4acde50
Show file tree

Hide file tree

Showing 4 changed files with 30 additions and 29 deletions.
diff --git a/macrel/main.py b/macrel/main.py
@@ -287,15 +287,16 @@ def do_predict(args, tdir):
 def do_density(args, clen, prediction):
     tpred = prediction.reset_index()
     tpred['contig'] = tpred['index'].apply(lambda x: '_'.join(x.split('_')[:-1]))
-    tpred = tpred[tpred['AMP_probability'] > 0.5]
+    tpred = tpred.query('AMP_probability > 0.5')
     tpred = tpred.groupby('contig').agg('size')
     tpred = tpred.reset_index()
     tpred = tpred.rename({0: 'AMPs'}, axis=1)
     clen = clen.merge(on='contig', right=tpred, how='outer').fillna(0)
-    clen[clen.columns[1:]] = clen[clen.columns[1:]].astype(int)   
+    clen[clen.columns[1:]] = clen[clen.columns[1:]].astype(int)
     ofile = path.join(args.output, args.outtag + '.percontigs.gz')
     sample = clen.set_index('contig').sum(axis=0).tolist()
     sample_density = sample[-1] * 1e6 / sample[0]
+    clen.sort_values('contig', inplace=True)
     with open_output(ofile, mode='wb') as raw_out:
         with gzip.open(raw_out, 'wt') as out:
             from .macrel_version import __version__

diff --git a/tests/contigs/expected.percontigs b/tests/contigs/expected.percontigs
@@ -1,27 +1,27 @@
 # Prediction from macrel v1.3.0
 # Macrel calculated for the sample a density of 45.062 AMPs / Mbp.
 contig	length	ORFs	smORFs	AMPs
-scaffold2530_2_MH0058	717	2	2	0
-scaffold75334_1_MH0058	3424	1	1	1
-scaffold54112_5_MH0058	728	1	1	0
-scaffold24504_2_MH0058	505	1	1	0
-scaffold95393_2_MH0058	995	2	1	0
-scaffold8449_1_MH0058	1037	2	1	0
+C4060843_1_MH0058	518	1	1	0
+C4067509_1_MH0058	534	1	1	0
+C4177507_1_MH0058	1322	2	1	0
+C4193751_1_MH0058	1820	3	1	0
 scaffold10455_3_MH0058	808	3	3	0
-scaffold98598_5_MH0058	3426	4	2	0
-scaffold76045_5_MH0058	960	3	3	0
 scaffold106190_2_MH0058	688	1	1	0
-C4067509_1_MH0058	534	1	1	0
-scaffold90770_1_MH0058	1031	2	2	0
-scaffold33693_17_MH0058	3481	2	1	1
-scaffold77554_3_MH0058	6086	9	4	0
-scaffold34596_7_MH0058	1345	1	0	0
-scaffold75223_9_MH0058	3597	2	2	0
-scaffold30291_4_MH0058	824	2	1	0
 scaffold107406_2_MH0058	1401	2	1	0
 scaffold16564_1_MH0058	992	2	2	0
-scaffold7019_2_MH0058	5218	6	3	0
-C4060843_1_MH0058	518	1	1	0
 scaffold20234_2_MH0058	2926	4	1	0
-C4193751_1_MH0058	1820	3	1	0
-C4177507_1_MH0058	1322	2	1	0
+scaffold24504_2_MH0058	505	1	1	0
+scaffold2530_2_MH0058	717	2	2	0
+scaffold30291_4_MH0058	824	2	1	0
+scaffold33693_17_MH0058	3481	2	1	1
+scaffold34596_7_MH0058	1345	1	0	0
+scaffold54112_5_MH0058	728	1	1	0
+scaffold7019_2_MH0058	5218	6	3	0
+scaffold75223_9_MH0058	3597	2	2	0
+scaffold75334_1_MH0058	3424	1	1	1
+scaffold76045_5_MH0058	960	3	3	0
+scaffold77554_3_MH0058	6086	9	4	0
+scaffold8449_1_MH0058	1037	2	1	0
+scaffold90770_1_MH0058	1031	2	2	0
+scaffold95393_2_MH0058	995	2	1	0
+scaffold98598_5_MH0058	3426	4	2	0
diff --git a/tests/reads.se/expected.percontigs b/tests/reads.se/expected.percontigs
@@ -3,12 +3,12 @@
 contig	length	ORFs	smORFs	AMPs
 k47_0	3379	4	2	0
 k47_1	6046	9	4	0
-k47_3	5202	6	3	0
-k47_7	3483	2	2	0
-k47_8	2877	3	1	0
 k47_10	3374	1	1	1
 k47_11	3415	2	1	1
 k47_12	1307	1	0	0
 k47_16	1790	3	1	0
 k47_17	1376	2	1	0
 k47_24	1228	2	1	0
+k47_3	5202	6	3	0
+k47_7	3483	2	2	0
+k47_8	2877	3	1	0
diff --git a/tests/reads/expected.percontigs b/tests/reads/expected.percontigs
@@ -1,15 +1,15 @@
 # Prediction from macrel v1.3.0
 # Macrel calculated for the sample a density of 57.627 AMPs / Mbp.
 contig	length	ORFs	smORFs	AMPs
-k77_3	1270	1	0	0
-k77_5	5202	6	3	0
-k77_6	3527	2	2	0
-k77_7	6058	9	4	0
-k77_8	1009	2	1	0
 k77_11	1303	2	1	0
 k77_12	3374	1	1	1
 k77_13	2920	4	1	0
 k77_15	3457	2	1	1
 k77_16	1790	3	1	0
 k77_20	1380	2	1	0
 k77_23	3416	4	2	0
+k77_3	1270	1	0	0
+k77_5	5202	6	3	0
+k77_6	3527	2	2	0
+k77_7	6058	9	4	0
+k77_8	1009	2	1	0