From ffeb3a61f770a78ad5b47e5587bae5a9bcb2f77d Mon Sep 17 00:00:00 2001
From: Alexander Kabui
Date: Tue, 8 Jun 2021 01:40:47 +0300
Subject: initial commit for queries perfomance tests

---
 tests/performance/__init__.py   |   0
 tests/performance/test_query.py | 119 ++++++++++++++++++++++++++++++++++++++++
 2 files changed, 119 insertions(+)
 create mode 100644 tests/performance/__init__.py
 create mode 100644 tests/performance/test_query.py

diff --git a/tests/performance/__init__.py b/tests/performance/__init__.py
new file mode 100644
index 0000000..e69de29
diff --git a/tests/performance/test_query.py b/tests/performance/test_query.py
new file mode 100644
index 0000000..c0a3210
--- /dev/null
+++ b/tests/performance/test_query.py
@@ -0,0 +1,119 @@
+"""module contains performance tests for queries"""
+
+import time
+import sys
+
+from inspect import getmembers
+from inspect import isfunction
+
+from functools import wraps
+from gn3.db_utils import database_connector
+
+
+def timer(func):
+    """time function"""
+    @wraps(func)
+    def wrapper_time(*args, **kwargs):
+        """time wrapper"""
+        start_time = time.perf_counter()
+        results = func(*args, **kwargs)
+        end_time = time.perf_counter()
+        run_time = end_time - start_time
+        print(f"the time taken is {run_time:.3f} seconds")
+        return results
+
+    return wrapper_time
+
+
+def query_executor(query, fetch_all=True):
+    """function to execute a query"""
+    conn, _ = database_connector()
+
+    with conn:
+        cursor = conn.cursor()
+        cursor.execute(query)
+
+        if fetch_all:
+            return cursor.fetchall()
+        return cursor.fetchone()
+
+
+def fetch_probeset_query(dataset_name):
+    """contains queries for datasets"""
+
+    query = """SELECT * from ProbeSetData
+            where StrainID in (4, 5, 6, 7, 8, 9, 10, 11, 12,
+            14, 15, 17, 18, 19, 20, 21, 22, 24, 25, 26, 28,
+             29, 30, 31, 35, 36, 37, 39, 98, 99, 100, 103,
+            487, 105, 106, 110, 115,116, 117, 118, 119, 
+            120, 919, 147, 121, 40, 41, 124, 125, 128, 135,
+            129, 130, 131, 132, 134, 138, 139, 140, 141, 142, 
+            144, 145, 148, 149, 920, 922, 2, 3, 1, 1100)
+            and id in (SELECT ProbeSetXRef.DataId
+            FROM (ProbeSet, ProbeSetXRef, ProbeSetFreeze)
+            WHERE ProbeSetXRef.ProbeSetFreezeId = ProbeSetFreeze.Id
+            and ProbeSetFreeze.Name = '{}'
+            and ProbeSet.Id = ProbeSetXRef.ProbeSetId)""".format(dataset_name)
+
+    return query
+
+
+@timer
+def perf_simple_query():
+    """initial simple query test"""
+
+    query = """select * from ProbeSetData limit 1"""
+
+    _results = query_executor(query)
+
+    return {}
+
+
+@timer
+def perf_hc_m2_dataset():
+    """test the default dataset HC_M2_0606_P"""
+
+    query = fetch_probeset_query("HC_M2_0606_P")
+
+    _results = query_executor(query)
+
+    return {}
+
+
+@timer
+def perf_umutaffyexon_dataset():
+    """largest dataset in gn"""
+
+    query = fetch_probeset_query("UMUTAffyExon_0209_RMA")
+    _results = query_executor(query)
+    return {}
+
+
+def fetch_perf_functions():
+    """function to filter all functions strwith perf_"""
+    name_func_dict = {name: obj for name, obj in
+                      getmembers(sys.modules[__name__], isfunction)if isfunction(
+                          obj) and obj.__module__ == __name__ and name.startswith('perf_')}
+
+    return name_func_dict
+
+
+def fetch_cmd_args():
+    """function to fetch cmd args"""
+    cmd_args = sys.argv[1:]
+
+    name_func_dict = fetch_perf_functions()
+
+    if len(cmd_args) > 0:
+        callables = [func_call for name,
+                     func_call in name_func_dict.items() if name in cmd_args]
+
+        return callables
+
+    return list(name_func_dict.values())
+
+
+if __name__ == '__main__':
+    func_list = fetch_cmd_args()
+    for call_func in func_list:
+        call_func()
-- 
cgit v1.2.3


From d15cbf953bd3b3e9260e88d5663a1cf20414d776 Mon Sep 17 00:00:00 2001
From: Alexander Kabui
Date: Tue, 8 Jun 2021 03:00:35 +0300
Subject: add types and dataset names

---
 tests/performance/test_query.py | 25 ++++++++++++++++---------
 1 file changed, 16 insertions(+), 9 deletions(-)

diff --git a/tests/performance/test_query.py b/tests/performance/test_query.py
index c0a3210..222e8d6 100644
--- a/tests/performance/test_query.py
+++ b/tests/performance/test_query.py
@@ -6,6 +6,7 @@ import sys
 from inspect import getmembers
 from inspect import isfunction
 
+from typing import Optional
 from functools import wraps
 from gn3.db_utils import database_connector
 
@@ -25,9 +26,12 @@ def timer(func):
     return wrapper_time
 
 
-def query_executor(query, fetch_all=True):
+def query_executor(query: str,
+                   dataset_name: Optional[str] = "dataset_name",
+                   fetch_all: bool = True):
     """function to execute a query"""
     conn, _ = database_connector()
+    print(f"Performance tests for {dataset_name}")
 
     with conn:
         cursor = conn.cursor()
@@ -38,7 +42,7 @@ def query_executor(query, fetch_all=True):
         return cursor.fetchone()
 
 
-def fetch_probeset_query(dataset_name):
+def fetch_probeset_query(dataset_name: str):
     """contains queries for datasets"""
 
     query = """SELECT * from ProbeSetData
@@ -75,7 +79,7 @@ def perf_hc_m2_dataset():
 
     query = fetch_probeset_query("HC_M2_0606_P")
 
-    _results = query_executor(query)
+    _results = query_executor(query, "HC_M2_0606_P")
 
     return {}
 
@@ -85,21 +89,24 @@ def perf_umutaffyexon_dataset():
     """largest dataset in gn"""
 
     query = fetch_probeset_query("UMUTAffyExon_0209_RMA")
-    _results = query_executor(query)
+    _results = query_executor(query, "UMUTAffyExon_0209_RMA")
     return {}
 
 
 def fetch_perf_functions():
     """function to filter all functions strwith perf_"""
-    name_func_dict = {name: obj for name, obj in
+    name_func_dict = {name: func_obj for name, func_obj in
                       getmembers(sys.modules[__name__], isfunction)if isfunction(
-                          obj) and obj.__module__ == __name__ and name.startswith('perf_')}
+                          func_obj)
+                      and func_obj.__module__ == __name__ and name.startswith('perf_')}
 
     return name_func_dict
 
 
 def fetch_cmd_args():
-    """function to fetch cmd args"""
+    """function to fetch cmd args\
+    for example python file.py perf_hc_m2_dataset\
+    output [perf_hc_m2_dataset obj]"""
     cmd_args = sys.argv[1:]
 
     name_func_dict = fetch_perf_functions()
@@ -115,5 +122,5 @@ def fetch_cmd_args():
 
 if __name__ == '__main__':
     func_list = fetch_cmd_args()
-    for call_func in func_list:
-        call_func()
+    for func in func_list:
+        func()
-- 
cgit v1.2.3


From 348f48b68eb029f9b3f74c9b14de508f643ce3fe Mon Sep 17 00:00:00 2001
From: Alexander Kabui
Date: Tue, 8 Jun 2021 03:40:18 +0300
Subject: add tests to setup packages

---
 setup.py | 1 +
 1 file changed, 1 insertion(+)

diff --git a/setup.py b/setup.py
index e43ab0a..3f0922b 100644
--- a/setup.py
+++ b/setup.py
@@ -30,6 +30,7 @@ setup(author='Bonface M. K.',
           'gn3.api',
           'gn3.computations',
           'gn3.db',
+          'tests'
       ],
       url='https://github.com/genenetwork/genenetwork3',
       version='0.1')
-- 
cgit v1.2.3


From 2431d8538625da97b07571c61f0d0dd29c5d880a Mon Sep 17 00:00:00 2001
From: Alexander Kabui
Date: Tue, 8 Jun 2021 03:48:04 +0300
Subject: pep8 formatting

---
 tests/performance/test_query.py | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/tests/performance/test_query.py b/tests/performance/test_query.py
index 222e8d6..2d05d26 100644
--- a/tests/performance/test_query.py
+++ b/tests/performance/test_query.py
@@ -31,7 +31,7 @@ def query_executor(query: str,
                    fetch_all: bool = True):
     """function to execute a query"""
     conn, _ = database_connector()
-    print(f"Performance tests for {dataset_name}")
+    print(f"Performance test for {dataset_name}")
 
     with conn:
         cursor = conn.cursor()
@@ -122,5 +122,5 @@ def fetch_cmd_args():
 
 if __name__ == '__main__':
     func_list = fetch_cmd_args()
-    for func in func_list:
-        func()
+    for func_obj in func_list:
+        func_obj()
-- 
cgit v1.2.3


From 72b3396ac8ccb4023122bd66d93c7cc63b4b6e6f Mon Sep 17 00:00:00 2001
From: Alexander Kabui
Date: Wed, 9 Jun 2021 20:23:58 +0300
Subject: rename perf query file

---
 tests/performance/test_query.py | 126 ----------------------------------------
 1 file changed, 126 deletions(-)
 delete mode 100644 tests/performance/test_query.py

diff --git a/tests/performance/test_query.py b/tests/performance/test_query.py
deleted file mode 100644
index 2d05d26..0000000
--- a/tests/performance/test_query.py
+++ /dev/null
@@ -1,126 +0,0 @@
-"""module contains performance tests for queries"""
-
-import time
-import sys
-
-from inspect import getmembers
-from inspect import isfunction
-
-from typing import Optional
-from functools import wraps
-from gn3.db_utils import database_connector
-
-
-def timer(func):
-    """time function"""
-    @wraps(func)
-    def wrapper_time(*args, **kwargs):
-        """time wrapper"""
-        start_time = time.perf_counter()
-        results = func(*args, **kwargs)
-        end_time = time.perf_counter()
-        run_time = end_time - start_time
-        print(f"the time taken is {run_time:.3f} seconds")
-        return results
-
-    return wrapper_time
-
-
-def query_executor(query: str,
-                   dataset_name: Optional[str] = "dataset_name",
-                   fetch_all: bool = True):
-    """function to execute a query"""
-    conn, _ = database_connector()
-    print(f"Performance test for {dataset_name}")
-
-    with conn:
-        cursor = conn.cursor()
-        cursor.execute(query)
-
-        if fetch_all:
-            return cursor.fetchall()
-        return cursor.fetchone()
-
-
-def fetch_probeset_query(dataset_name: str):
-    """contains queries for datasets"""
-
-    query = """SELECT * from ProbeSetData
-            where StrainID in (4, 5, 6, 7, 8, 9, 10, 11, 12,
-            14, 15, 17, 18, 19, 20, 21, 22, 24, 25, 26, 28,
-             29, 30, 31, 35, 36, 37, 39, 98, 99, 100, 103,
-            487, 105, 106, 110, 115,116, 117, 118, 119, 
-            120, 919, 147, 121, 40, 41, 124, 125, 128, 135,
-            129, 130, 131, 132, 134, 138, 139, 140, 141, 142, 
-            144, 145, 148, 149, 920, 922, 2, 3, 1, 1100)
-            and id in (SELECT ProbeSetXRef.DataId
-            FROM (ProbeSet, ProbeSetXRef, ProbeSetFreeze)
-            WHERE ProbeSetXRef.ProbeSetFreezeId = ProbeSetFreeze.Id
-            and ProbeSetFreeze.Name = '{}'
-            and ProbeSet.Id = ProbeSetXRef.ProbeSetId)""".format(dataset_name)
-
-    return query
-
-
-@timer
-def perf_simple_query():
-    """initial simple query test"""
-
-    query = """select * from ProbeSetData limit 1"""
-
-    _results = query_executor(query)
-
-    return {}
-
-
-@timer
-def perf_hc_m2_dataset():
-    """test the default dataset HC_M2_0606_P"""
-
-    query = fetch_probeset_query("HC_M2_0606_P")
-
-    _results = query_executor(query, "HC_M2_0606_P")
-
-    return {}
-
-
-@timer
-def perf_umutaffyexon_dataset():
-    """largest dataset in gn"""
-
-    query = fetch_probeset_query("UMUTAffyExon_0209_RMA")
-    _results = query_executor(query, "UMUTAffyExon_0209_RMA")
-    return {}
-
-
-def fetch_perf_functions():
-    """function to filter all functions strwith perf_"""
-    name_func_dict = {name: func_obj for name, func_obj in
-                      getmembers(sys.modules[__name__], isfunction)if isfunction(
-                          func_obj)
-                      and func_obj.__module__ == __name__ and name.startswith('perf_')}
-
-    return name_func_dict
-
-
-def fetch_cmd_args():
-    """function to fetch cmd args\
-    for example python file.py perf_hc_m2_dataset\
-    output [perf_hc_m2_dataset obj]"""
-    cmd_args = sys.argv[1:]
-
-    name_func_dict = fetch_perf_functions()
-
-    if len(cmd_args) > 0:
-        callables = [func_call for name,
-                     func_call in name_func_dict.items() if name in cmd_args]
-
-        return callables
-
-    return list(name_func_dict.values())
-
-
-if __name__ == '__main__':
-    func_list = fetch_cmd_args()
-    for func_obj in func_list:
-        func_obj()
-- 
cgit v1.2.3


From 16fc8e6be584299eb839bf1be54d7ee472d2321b Mon Sep 17 00:00:00 2001
From: Alexander Kabui
Date: Wed, 9 Jun 2021 20:24:36 +0300
Subject: refactor perf query functions

---
 tests/performance/perf_query.py | 113 ++++++++++++++++++++++++++++++++++++++++
 1 file changed, 113 insertions(+)
 create mode 100644 tests/performance/perf_query.py

diff --git a/tests/performance/perf_query.py b/tests/performance/perf_query.py
new file mode 100644
index 0000000..594b9ea
--- /dev/null
+++ b/tests/performance/perf_query.py
@@ -0,0 +1,113 @@
+"""module contains performance tests for queries"""
+
+import time
+import sys
+
+from inspect import getmembers
+from inspect import isfunction
+
+from typing import Optional
+from functools import wraps
+from gn3.db_utils import database_connector
+
+
+def timer(func):
+    """time function"""
+    @wraps(func)
+    def wrapper_time(*args, **kwargs):
+        """time wrapper"""
+        start_time = time.perf_counter()
+        results = func(*args, **kwargs)
+        end_time = time.perf_counter()
+        run_time = end_time - start_time
+        print(f"the time taken is {run_time:.3f} seconds")
+        return results
+
+    return wrapper_time
+
+
+def query_executor(query: str,
+                   dataset_name: Optional[str] = "dataset_name",
+                   fetch_all: bool = True):
+    """function to execute a query"""
+    conn, _ = database_connector()
+
+    with conn:
+        cursor = conn.cursor()
+        cursor.execute(query)
+
+        if fetch_all:
+            return cursor.fetchall()
+        return cursor.fetchone()
+
+
+def fetch_probeset_query(dataset_name: str):
+    """contains queries for datasets"""
+
+    query = """SELECT * from ProbeSetData
+            where StrainID in (4, 5, 6, 7, 8, 9, 10, 11, 12,
+            14, 15, 17, 18, 19, 20, 21, 22, 24, 25, 26, 28,
+             29, 30, 31, 35, 36, 37, 39, 98, 99, 100, 103,
+            487, 105, 106, 110, 115,116, 117, 118, 119, 
+            120, 919, 147, 121, 40, 41, 124, 125, 128, 135,
+            129, 130, 131, 132, 134, 138, 139, 140, 141, 142, 
+            144, 145, 148, 149, 920, 922, 2, 3, 1, 1100)
+            and id in (SELECT ProbeSetXRef.DataId
+            FROM (ProbeSet, ProbeSetXRef, ProbeSetFreeze)
+            WHERE ProbeSetXRef.ProbeSetFreezeId = ProbeSetFreeze.Id
+            and ProbeSetFreeze.Name = '{}'
+            and ProbeSet.Id = ProbeSetXRef.ProbeSetId)""".format(dataset_name)
+
+    return query
+
+
+@timer
+def perf_hc_m2_dataset():
+    """test the default dataset HC_M2_0606_P"""
+
+    dataset_name = "HC_M2_0606_P"
+    print(f"Performance test for {dataset_name}")
+
+    query_executor(fetch_probeset_query(dataset_name=dataset_name),
+                   dataset_name=dataset_name)
+
+
+@timer
+def perf_umutaffyexon_dataset():
+    """largest dataset in gn"""
+
+    dataset_name = "UMUTAffyExon_0209_RMA"
+    print(f"Performance test for {dataset_name}")
+    query_executor(fetch_probeset_query(dataset_name=dataset_name),
+                   dataset_name=dataset_name)
+
+
+def fetch_perf_functions():
+    """function to filter all functions strwith perf_"""
+    name_func_dict = {name: func_obj for name, func_obj in
+                      getmembers(sys.modules[__name__], isfunction)if isfunction(
+                          func_obj)
+                      and func_obj.__module__ == __name__ and name.startswith('perf_')}
+
+    return name_func_dict
+
+
+def fetch_cmd_args():
+    """function to fetch cmd args"""
+    cmd_args = sys.argv[1:]
+
+    name_func_dict = fetch_perf_functions()
+
+    if len(cmd_args) > 0:
+        callables = [func_call for name,
+                     func_call in name_func_dict.items() if name in cmd_args]
+
+        return callables
+
+    return list(name_func_dict.values())
+
+
+if __name__ == '__main__':
+    func_list = fetch_cmd_args()
+    for func_obj in func_list:
+        func_obj()
-- 
cgit v1.2.3


From c361cc5c16914e43689db8b7f723c32986e4cc20 Mon Sep 17 00:00:00 2001
From: Alexander Kabui
Date: Wed, 9 Jun 2021 20:27:20 +0300
Subject: remove ununsed variables

---
 tests/performance/perf_query.py | 1 -
 1 file changed, 1 deletion(-)

diff --git a/tests/performance/perf_query.py b/tests/performance/perf_query.py
index 594b9ea..93bb3b2 100644
--- a/tests/performance/perf_query.py
+++ b/tests/performance/perf_query.py
@@ -27,7 +27,6 @@ def timer(func):
 
 
 def query_executor(query: str,
-                   dataset_name: Optional[str] = "dataset_name",
                    fetch_all: bool = True):
     """function to execute a query"""
     conn, _ = database_connector()
-- 
cgit v1.2.3


From 9d6af6049fa73c6aae4d224245d319e87bccbd6a Mon Sep 17 00:00:00 2001
From: Alexander Kabui
Date: Wed, 9 Jun 2021 20:31:43 +0300
Subject: minor fixes

---
 tests/performance/perf_query.py | 7 ++-----
 1 file changed, 2 insertions(+), 5 deletions(-)

diff --git a/tests/performance/perf_query.py b/tests/performance/perf_query.py
index 93bb3b2..12cb944 100644
--- a/tests/performance/perf_query.py
+++ b/tests/performance/perf_query.py
@@ -6,7 +6,6 @@ import sys
 from inspect import getmembers
 from inspect import isfunction
 
-from typing import Optional
 from functools import wraps
 from gn3.db_utils import database_connector
 
@@ -67,8 +66,7 @@ def perf_hc_m2_dataset():
     dataset_name = "HC_M2_0606_P"
     print(f"Performance test for {dataset_name}")
 
-    query_executor(fetch_probeset_query(dataset_name=dataset_name),
-                   dataset_name=dataset_name)
+    query_executor(fetch_probeset_query(dataset_name=dataset_name))
 
 
 @timer
@@ -77,8 +75,7 @@ def perf_umutaffyexon_dataset():
 
     dataset_name = "UMUTAffyExon_0209_RMA"
     print(f"Performance test for {dataset_name}")
-    query_executor(fetch_probeset_query(dataset_name=dataset_name),
-                   dataset_name=dataset_name)
+    query_executor(fetch_probeset_query(dataset_name=dataset_name))
 
 
 def fetch_perf_functions():
-- 
cgit v1.2.3