def loadDemoTables():
'''
Load the tables used in the Demo, if they don't exist.
'''
dbconn = DBConnect()
conn_dict = dbconn.getConnectionString()
load_tbl_stmt = '''psql -h {hostname} -p {port} -U {username} -d {database} -f '''
load_tbl_stmt = load_tbl_stmt.format(username=conn_dict['username'],
hostname=conn_dict['hostname'],
database=conn_dict['database'],
port=conn_dict['port'])
this_dir = os.path.dirname(os.path.abspath(__file__))
for fl in os.listdir(os.path.join(this_dir, 'data/')):
full_path = os.path.join(os.path.join(this_dir, 'data'), fl)
if (fl.endswith('.sql')):
#If the demo table does not exists in the database already, create it using the provided sql files
if (not __isTableExists__(fl[:-len('.sql')], dbconn)):
logging.info('cmd:{0}'.format(load_tbl_stmt + ' ' + full_path))
cmd = load_tbl_stmt + ' ' + full_path
os.system(cmd)
logging.info('Loading demo tables complete')
def loadDemoTables():
'''
Load the tables used in the Demo, if they don't exist.
'''
dbconn = DBConnect()
conn_dict = dbconn.getConnectionString()
load_tbl_stmt = '''psql -h {hostname} -p {port} -U {username} -d {database} -f '''
load_tbl_stmt = load_tbl_stmt.format( username=conn_dict['username'],
hostname=conn_dict['hostname'],
database=conn_dict['database'],
port=conn_dict['port']
)
this_dir = os.path.dirname(os.path.abspath(__file__))
for fl in os.listdir(os.path.join(this_dir,'data/')):
full_path = os.path.join(os.path.join(this_dir,'data'),fl)
if(fl.endswith('.sql')):
#If the demo table does not exists in the database already, create it using the provided sql files
if(not __isTableExists__(fl[:-len('.sql')],dbconn)):
logging.info('cmd:{0}'.format(load_tbl_stmt+' '+full_path))
cmd = load_tbl_stmt+ ' '+full_path
os.system(cmd)
logging.info('Loading demo tables complete')
def conn_test():
'''
Test the connection by displaying rows from a table
'''
conn = DBConnect()
cursor = conn.getCursor(True)
cursor.executeQuery('select * from wine_training_set')
conn.printTable(cursor)
def pyMADlibDemo():
'''
Demonstrate building Linear Regression and Logistic Regression Models using MADlib
'''
conn = DBConnect(madlib_schema='madlib_v05')
#1) Linear Regression
linearRegressionDemo(conn)
#2) Logistic Regression
logisticRegDemo(conn)
#3) SVM Regression
svmDemo(conn)
#4) KMeans
kmeansDemo(conn)
#5) PLDA
pldaDemo(conn)
drop table if exists {output_table} cascade;
create table {output_table} as
(
select array[{list_of_indep_cols}] as {indep_cols_arr_name}
from {table_name}
);
'''
convert_to_arr_stmt = convert_to_arr_stmt.format(**data_dict)
conn.executeQuery(convert_to_arr_stmt)
return output_table, indep_cols_arr_name
if (__name__ == '__main__'):
from pymadlib import DBConnect
conn = DBConnect()
output_table, indep, dep, cols_distinct_vals = pivotCategoricalColumns(
conn, 'cuse_dat', ['1', 'age', 'education', 'wantsmore', 'notusing'],
'yesusing')
print 'output table :', output_table
print 'output independent columns :', indep
print 'dependent col :', dep
#Verify if the input has all numeric columns, the input table is returned unchanged.
output_table, indep, dep, cols_distinct_vals = pivotCategoricalColumns(
conn, 'wine_training_set',
['1', 'alcohol', 'proline', 'hue', 'color_intensity', 'flavanoids'],
'quality')
print 'output table :', output_table
print 'output independent columns :', indep
print 'dependent col :', dep
