aboutsummaryrefslogtreecommitdiff
diff options
context:
space:
mode:
authorzsloan2020-11-13 15:25:32 -0600
committerzsloan2020-11-13 15:25:32 -0600
commit76d5b24b3d2e0b5e6be5b1cf4c125a9a812ff57a (patch)
treea32c7ff1a89892f5343fadaa4340522d371e214b
parent975fb0a19feafaf0f2c9a9dbcd88b96a3439196a (diff)
downloadgenenetwork2-76d5b24b3d2e0b5e6be5b1cf4c125a9a812ff57a.tar.gz
Switched to using boolean casting to check if various trait attributes
exist
-rw-r--r--wqflask/wqflask/correlation/show_corr_results.py59
1 files changed, 29 insertions, 30 deletions
diff --git a/wqflask/wqflask/correlation/show_corr_results.py b/wqflask/wqflask/correlation/show_corr_results.py
index 17d3de54..87dd7597 100644
--- a/wqflask/wqflask/correlation/show_corr_results.py
+++ b/wqflask/wqflask/correlation/show_corr_results.py
@@ -494,61 +494,60 @@ def generate_corr_json(corr_results, this_trait, dataset, target_dataset, for_ap
results_dict['mean'] = "N/A"
results_dict['lrs_score'] = "N/A"
results_dict['additive'] = "N/A"
- if trait.description_display:
+ if bool(trait.description_display):
results_dict['description'] = trait.description_display
- if trait.mean and trait.mean != "":
+ if bool(trait.mean):
results_dict['mean'] = float(trait.mean)
if trait.LRS_score_repr != "N/A":
results_dict['lrs_score'] = "%.1f" % float(trait.LRS_score_repr)
results_dict['lrs_location'] = trait.LRS_location_repr
- if trait.additive and trait.additive != "":
+ if bool(trait.additive):
results_dict['additive'] = "%0.3f" % float(trait.additive)
results_dict['sample_r'] = "%0.3f" % float(trait.sample_r)
results_dict['num_overlap'] = trait.num_overlap
results_dict['sample_p'] = "%0.3e" % float(trait.sample_p)
- if trait.lit_corr == "" or trait.lit_corr == 0:
- results_dict['lit_corr'] = "--"
- else:
+ results_dict['lit_corr'] = "--"
+ results_dict['tissue_corr'] = "--"
+ results_dict['tissue_pvalue'] = "--"
+ if bool(trait.lit_corr):
results_dict['lit_corr'] = "%0.3f" % float(trait.lit_corr)
- if trait.tissue_corr == "" or trait.tissue_corr == 0:
- results_dict['tissue_corr'] = "--"
- results_dict['tissue_pvalue'] = "--"
- else:
+ if bool(trait.tissue_corr):
results_dict['tissue_corr'] = "%0.3f" % float(trait.tissue_corr)
results_dict['tissue_pvalue'] = "%0.3e" % float(trait.tissue_pvalue)
elif target_dataset.type == "Publish":
results_dict['abbreviation_display'] = "N/A"
- if trait.abbreviation:
- results_dict['abbreviation_display'] = trait.abbreviation
results_dict['description'] = "N/A"
- if trait.description_display:
- results_dict['description'] = trait.description_display
- results_dict['authors'] = trait.authors
- authors_list = trait.authors.split(',')
- if len(authors_list) > 6:
- results_dict['authors_display'] = ", ".join(authors_list[:6]) + ", et al."
+ results_dict['authors_display'] = "N/A"
+ results_dict['additive'] = "N/A"
+ if for_api:
+ results_dict['pubmed_id'] = "N/A"
+ results_dict['year'] = "N/A"
else:
- results_dict['authors_display'] = trait.authors
- if trait.pubmed_id:
+ results_dict['pubmed_link'] = "N/A"
+ results_dict['pubmed_text'] = "N/A"
+
+ if bool(trait.abbreviation):
+ results_dict['abbreviation_display'] = trait.abbreviation
+ if bool(trait.description_display):
+ results_dict['description'] = trait.description_display
+ if bool(trait.authors):
+ authors_list = trait.authors.split(',')
+ if len(authors_list) > 6:
+ results_dict['authors_display'] = ", ".join(authors_list[:6]) + ", et al."
+ else:
+ results_dict['authors_display'] = trait.authors
+ if bool(trait.pubmed_id):
if for_api:
results_dict['pubmed_id'] = trait.pubmed_id
results_dict['year'] = trait.pubmed_text
else:
results_dict['pubmed_link'] = trait.pubmed_link
results_dict['pubmed_text'] = trait.pubmed_text
- else:
- if for_api:
- results_dict['pubmed_id'] = "N/A"
- results_dict['year'] = "N/A"
- else:
- results_dict['pubmed_link'] = "N/A"
- results_dict['pubmed_text'] = "N/A"
+
results_dict['lrs_score'] = trait.LRS_score_repr
results_dict['lrs_location'] = trait.LRS_location_repr
- if trait.additive and trait.additive != "":
+ if bool(trait.additive):
results_dict['additive'] = "%0.3f" % float(trait.additive)
- else:
- results_dict['additive'] = "N/A"
results_dict['sample_r'] = "%0.3f" % trait.sample_r
results_dict['num_overlap'] = trait.num_overlap
results_dict['sample_p'] = "%0.3e" % float(trait.sample_p)