def load(cls, fname=None):
'''Loads a config file from the specified file name, `fname`. If `fname` is
not specified, uses `DEFAULT_PATH`.
Args:
fname, str: The file name.
Throws:
ParseError: If the config file was malformatted.
Returns:
The `Config` object found in fname.
'''
fname = fname or DEFAULT_PATH
with open(fname) as f:
file_contents = "".join(f.readlines())
dbg('got', file_contents)
return Config(json.loads(file_contents))
"-p",
"--path",
required=True,
help="Path to the directory containing fasta and binding.txt files")
ap.add_argument("-a", "--allelle", required=True, help="Name of allelle")
ap.add_argument("-l",
"--alignment-file-name",
required=True,
help="Name of alignment file")
ap.add_argument("-b",
"--binding-file-name",
required=True,
help="Name of binding file")
args = vars(ap.parse_args())
dbg('args["path"]')
dbg(args["path"])
dbg('args["allelle"]')
dbg(args["allelle"])
data_path_align = args["path"] + args["alignment_file_name"]
data_path_bind = args["path"] + args["binding_file_name"]
dbg('data_path_align')
dbg(data_path_align)
dbg('data_path_bind')
dbg(data_path_bind)
seq_array = pfun.read_seq_data(data_path=data_path_align)
"-p",
"--path",
required=True,
help="Path to the directory containing fasta and binding.txt files")
ap.add_argument("-a", "--allelle", required=True, help="Name of allelle")
ap.add_argument("-l",
"--alignment-file-name",
required=True,
help="Name of alignment file")
ap.add_argument("-b",
"--binding-file-name",
required=True,
help="Name of binding file")
args = vars(ap.parse_args())
dbg('args["path"]')
dbg(args["path"])
dbg('args["allelle"]')
dbg(args["allelle"])
data_path_align = args["path"] + args["alignment_file_name"]
data_path_bind = args["path"] + args["binding_file_name"]
model_prefix_str = 'TEST_catboost_model.allelle_._' + args["allelle"]
dbg('data_path_align')
dbg(data_path_align)
dbg('data_path_bind')
dbg(data_path_bind)
ap = argparse.ArgumentParser()
ap.add_argument(
"-pi",
"--path-input",
required=True,
help=
"Path to the directory containing face images and corresponding gan input vectors"
)
ap.add_argument(
"-po",
"--path-output",
required=True,
help="Path to the directory where numpy array files will be saved")
args = vars(ap.parse_args())
dbg('args["path_input"]')
dbg(args["path_input"])
fpi = args["path_input"]
dbg('args["path_output"]')
dbg(args["path_output"])
fpo = args["path_output"]
""" walk though path and get images"""
path_img = []
for (dirpath, dirnames, filenames) in os.walk(fpi):
for x in filenames:
tmp_ext = os.path.basename(x).split('.')[-1]
tmp_ext_lower = tmp_ext.lower()
#fig.savefig(var_name+'.png', bbox_inches='tight')
# define an empty list
auc_tmp = []
ppvn_tmp = []
auc_comp_list = []
ppvn_comp_list = []
folders = []
for (root, dirs, files
) in os.walk('../Reda-BG/MHC-I/generation-1/competition/MHCflurry/HPV/'):
dbg(root)
dbg(dirs)
dbg(files)
dbg('--------------------------------')
break
for item in dirs:
dbg('folder name')
dbg(item)
tmp_folder = './' + item + '/'
dbg('tmp_folder')
dbg(tmp_folder)
for (subroot, subdirs, subfiles) in os.walk(tmp_folder):
dbg('subfiles')
dbg(subfiles)
if ('auc.txt' in subfiles) and ('ppvn.txt' in subfiles):
def search_ct_gov(search_term='sars'):
max_res_pull = 100
# preprocess search string
# -- Replace space with '+'
search_term = search_term.replace(" ", "+")
# -- Replace & with '%26'
search_term = search_term.replace("&", "%26")
# Build Search URL
# -- example: search_url = 'https://clinicaltrials.gov/api/query/study_fields?fmt=CSV&expr=heart+attack&fields=NCTId,Condition,InterventionName'
search_url = 'https://clinicaltrials.gov/api/query/study_fields?fmt=CSV&expr=' + search_term + '&fields=NCTId,Condition,InterventionName,InterventionType,OfficialTitle'
# Execute get request
response = requests.get(search_url)
tmp_search_res = response.text
dbg(response.text)
# Splitting on newline delimiter
res_list = tmp_search_res.splitlines()
dbg(res_list)
# This code is written for clinicaltrials.gov API version "1.01.02"
# Ensure that implemented API verson has not changed
api_ver_str = res_list[0].replace('\"', '')
api_ver_num_str = api_ver_str.split(':')[1].strip()
dbg(api_ver_num_str)
api_ver_check = ("1.01.02" == api_ver_num_str)
dbg(api_ver_check)
# Extract total number of trials available for search query
num_trials_res_str = res_list[4].replace('\"', '')
num_trials_res_value = int(num_trials_res_str.split(':')[1].strip())
dbg(num_trials_res_value)
# Calculate number of get requests needed to grab all results
# -- max_res_pull is the max number of results that can be requested in each get request
# -- this limit is set by clinicaltrials.gov
total_num_pulls = (num_trials_res_value // max_res_pull) + 1
dbg(total_num_pulls)
all_search_results_list = []
all_search_results_str = ''
all_search_results_list.append(res_list[10].replace('\"', '').split(','))
all_search_results_str = all_search_results_str.join(res_list[10]) + '\n'
dbg(all_search_results_list)
dbg(all_search_results_str)
#&min_rnk=1198&max_rnk=1249&fmt=xml
for pull_idx in range(total_num_pulls):
min_rank = pull_idx * max_res_pull + 1
max_rank = (pull_idx + 1) * max_res_pull + 1
search_url = 'https://clinicaltrials.gov/api/query/study_fields?fmt=CSV&expr=' + search_term + '&fields=NCTId,Condition,InterventionName&min_rnk=' + str(
min_rank) + '&max_rnk=' + str(max_rank)
response = requests.get(search_url)
#tmp_search_res = response.text.replace('\"', '')
tmp_search_res = response.text
# Splitting on newline delimiter
res_list = tmp_search_res.splitlines()
res_list_comma_split = [x.split(',') for x in res_list]
dbg(res_list_comma_split[11:])
all_search_results_list.extend(res_list_comma_split[11:])
temp_results_str = '\n'.join(str(x) for x in res_list[11:])
all_search_results_str = all_search_results_str + temp_results_str + '\n'
dbg(all_search_results_list)
all_search_results_str_IO = StringIO(all_search_results_str)
df_search_results = pd.read_csv(all_search_results_str_IO)
dbg(df_search_results.head())
dbg(df_search_results.tail())
dbg(df_search_results.columns)
return df_search_results
def load_csv_ct_data(trial_file_list):
tot_df = pd.DataFrame()
for count, fn in enumerate(trial_file_list):
dbg('fn')
dbg(fn)
tmp_df = pd.read_csv(fn, encoding='ANSI')
tmp_df['class'] = count
dbg(tmp_df.head())
tot_df = pd.concat([tot_df, tmp_df.head(n=50)],
ignore_index=True,
sort=False)
#tot_df = pd.merge(tot_df, tmp_df, on='NCT Number')
dbg(tot_df)
dbg(tot_df.head())
dbg(tot_df.tail())
return tot_df