aboutsummaryrefslogtreecommitdiff
diff options
context:
space:
mode:
authorFrederick Muriuki Muriithi2022-07-28 10:21:58 +0300
committerFrederick Muriuki Muriithi2022-07-28 10:21:58 +0300
commit82922e5a052abf22b9d458ddebaf8ee523bd6f55 (patch)
tree74f4fdbce3d579f71b816d9d222f4c1e95383c4e
parenta08d91a234f700043d1d31164c7e2bacda4729da (diff)
downloadgenenetwork2-82922e5a052abf22b9d458ddebaf8ee523bd6f55.tar.gz
Use external process to compute sample correlations
-rw-r--r--wqflask/wqflask/correlation/correlation_gn3_api.py15
1 files changed, 6 insertions, 9 deletions
diff --git a/wqflask/wqflask/correlation/correlation_gn3_api.py b/wqflask/wqflask/correlation/correlation_gn3_api.py
index caa9b619..fe05635e 100644
--- a/wqflask/wqflask/correlation/correlation_gn3_api.py
+++ b/wqflask/wqflask/correlation/correlation_gn3_api.py
@@ -11,7 +11,7 @@ from base import data_set
from base.trait import create_trait
from base.trait import retrieve_sample_data
-from gn3.computations.correlations import compute_all_sample_correlation
+from gn3.commands import run_sample_corr_cmd
from gn3.computations.correlations import fast_compute_all_sample_correlation
from gn3.computations.correlations import map_shared_keys_to_values
from gn3.computations.correlations import compute_all_lit_correlation
@@ -108,9 +108,9 @@ def sample_for_trait_lists(corr_results, target_dataset,
(this_trait_data, target_dataset) = fetch_sample_data(
start_vars, this_trait, this_dataset, target_dataset)
- correlation_results = compute_all_sample_correlation(corr_method="pearson",
- this_trait=this_trait_data,
- target_dataset=target_dataset)
+ correlation_results = run_sample_corr_cmd(
+ corr_method="pearson", this_trait=this_trait_data,
+ target_dataset=target_dataset)
return correlation_results
@@ -211,9 +211,8 @@ def compute_correlation(start_vars, method="pearson", compute_all=False):
if corr_type == "sample":
(this_trait_data, target_dataset_data) = fetch_sample_data(
start_vars, this_trait, this_dataset, target_dataset)
-
- correlation_results = compute_all_sample_correlation(
- corr_method=method, this_trait=this_trait_data, target_dataset=target_dataset_data)
+ correlation_results = run_sample_corr_cmd(
+ method, this_trait_data, target_dataset_data)
elif corr_type == "tissue":
trait_symbol_dict = this_dataset.retrieve_genes("Symbol")
@@ -275,12 +274,10 @@ def compute_corr_for_top_results(start_vars,
target_dataset,
corr_type):
if corr_type != "tissue" and this_dataset.type == "ProbeSet" and target_dataset.type == "ProbeSet":
-
tissue_result = tissue_for_trait_lists(
correlation_results, this_dataset, this_trait)
if tissue_result:
-
correlation_results = merge_correlation_results(
correlation_results, tissue_result)