A keen python util so queries return a consistent panda dataframe with / without timeframes

gistfile1.txt

from datetime import date, timedelta
import datetime
import pandas as pd

import json

import yaml
from keen.client import KeenClient

def events():
	f = open('../config.yml')
	# use safe_load instead load
	configMap = yaml.safe_load(f)
	f.close()
	client = KeenClient(
		project_id=configMap['project_id'],
		read_key=configMap['read_key']
	)
	return client
def extract_timeinterval(keen_interval_extract):
   raw_data = []
   for item_row in keen_interval_extract:

       # Takes in a keen raw output, not a df
       this_item_row_start = datetime.datetime.strptime(item_row['timeframe']['start'], '%Y-%m-%dT%H:%M:%S.%fZ')
       if type(item_row['value']) == type(1):

           item_row['start']=this_item_row_start
           raw_data.append({'start': this_item_row_start, 'result': item_row['value'] })
       else:
           for this_element in item_row['value']:
                this_element['start']=this_item_row_start
                raw_data.append(this_element)
   return pd.DataFrame(raw_data)

def excluded_players(excluded_players_filename='../data/reference/excluded_players.json'):
    ## Define Player Exclusion Filters
    j_data=open(excluded_players_filename).read()
    j_data = json.loads(j_data)
    j_players_list = [j_data['excluded_players'][x]['player_id'] for x in range(0,len(j_data['excluded_players']))]
    j_players_list = [x.encode('UTF8') for x in j_players_list]
    exclude_players_filters = []
    for player in j_players_list:
       this_player_exclude = {"property_name": "player_id", "operator": "ne", "property_value": player }
       exclude_players_filters.append(this_player_exclude)
    return exclude_players_filters

def extract_variable(keen_interval_extract,variable_name,period_start):
   raw_df=pd.DataFrame(keen_interval_extract)
   raw_df['start']=raw_df['result'].map(lambda x: period_start)
   return raw_df


def keen_count(table,result_variable_name,period_start,period_finish,input_group_by_variables=[],selection_filters=[],interval=None):
    group_by_variables = input_group_by_variables
    group_df_variables = []
    for input_group_by_variable in input_group_by_variables:
      group_df_variables = group_df_variables + [input_group_by_variable]
    the_excluded_players=excluded_players()
    filters = selection_filters+the_excluded_players
    if interval == None:
        raw = (events().count(
        table,
        group_by=group_by_variables,
        filters=filters,
        timeframe={'start': period_start.strftime('%Y-%m-%dT%H:%M:%S.%fZ'), 'end': period_finish.strftime('%Y-%m-%dT%H:%M:%S.%fZ') },
        timezone='utc'
        ))
        if len(group_df_variables) == 0:
            raw_df = pd.DataFrame([{'start': period_start, 'result': raw}])
        else:
            raw_df=pd.DataFrame(raw)
        raw_df=raw_df.rename(columns={'result' : result_variable_name})
        return raw_df
    else:
        raw = (events().count(
        table,
        group_by=group_by_variables,
        filters=filters,
        timeframe={'start': period_start.strftime('%Y-%m-%dT%H:%M:%S.%fZ'), 'end': period_finish.strftime('%Y-%m-%dT%H:%M:%S.%fZ') },
        interval = interval,
        timezone='utc'))
        raw_df = pd.DataFrame(extract_timeinterval(raw))
        raw_df=raw_df.rename(columns={'result' : result_variable_name})

        if len(group_df_variables) == 0:
            return raw_df
        else:
          data_df = raw_df.groupby(group_df_variables+['start']).sum()
          return data_df


def keen_count_unique(table,target_variable,result_variable_name,period_start,period_finish,input_group_by_variables=[],selection_filters=[],interval=None):
    group_by_variables = input_group_by_variables
    group_df_variables = []
    for input_group_by_variable in input_group_by_variables:
      group_df_variables = group_df_variables + [input_group_by_variable]
    the_excluded_players=excluded_players()
    filters = selection_filters+the_excluded_players
    if interval == None:
        raw = (events().count_unique(
        table,
        group_by=group_by_variables,
        target_property=target_variable,
        filters=filters,
        timeframe={'start': period_start.strftime('%Y-%m-%dT%H:%M:%S.%fZ'), 'end': period_finish.strftime('%Y-%m-%dT%H:%M:%S.%fZ') },
        timezone='utc'
        ))
        if len(group_df_variables) == 0:
            raw_df = pd.DataFrame([{'start': period_start, 'result': raw}])
        else:
            raw_df=pd.DataFrame(raw)
        raw_df=raw_df.rename(columns={'result' : result_variable_name})
        return raw_df
    else:
        raw = (events().count_unique(
        table,
        target_property=target_variable,
        group_by=group_by_variables,
        filters=filters,
        timeframe={'start': period_start.strftime('%Y-%m-%dT%H:%M:%S.%fZ'), 'end': period_finish.strftime('%Y-%m-%dT%H:%M:%S.%fZ') },
        interval = interval,
        timezone='utc'))
        raw_df = pd.DataFrame(extract_timeinterval(raw))
        raw_df=raw_df.rename(columns={'result' : result_variable_name})

        if len(group_df_variables) == 0:
            return raw_df
        else:
          data_df = raw_df.groupby(group_df_variables+['start']).sum()
          return data_df

def keen_maximum(table,target_variable,result_variable_name,period_start,period_finish,input_group_by_variables=[],selection_filters=[],interval=None):
    group_by_variables = input_group_by_variables
    group_df_variables = []
    for input_group_by_variable in input_group_by_variables:
      group_df_variables = group_df_variables + [input_group_by_variable]
    the_excluded_players=excluded_players()
    filters = selection_filters+the_excluded_players
    if interval == None:
        raw = (events().maximum(
        table,
        group_by=group_by_variables,
        target_property=target_variable,
        filters=filters,
        timeframe={'start': period_start.strftime('%Y-%m-%dT%H:%M:%S.%fZ'), 'end': period_finish.strftime('%Y-%m-%dT%H:%M:%S.%fZ') },
        timezone='utc'
        ))
        if len(group_df_variables) == 0:
            raw_df = pd.DataFrame([{'start': period_start, 'result': raw}])
        else:
            raw_df=pd.DataFrame(raw)
        raw_df=raw_df.rename(columns={'result' : result_variable_name})
        return raw_df
    else:
        raw = (events().maximum(
        table,
        target_property=target_variable,
        group_by=group_by_variables,
        filters=filters,
        timeframe={'start': period_start.strftime('%Y-%m-%dT%H:%M:%S.%fZ'), 'end': period_finish.strftime('%Y-%m-%dT%H:%M:%S.%fZ') },
        interval = interval,
        timezone='utc'))
        raw_df = pd.DataFrame(extract_timeinterval(raw))
        raw_df=raw_df.rename(columns={'result' : result_variable_name})

        if len(group_df_variables) == 0:
            return raw_df
        else:
          data_df = raw_df.groupby(group_df_variables+['start']).sum()
          return data_df

def keen_minimum(table,target_variable,result_variable_name,period_start,period_finish,input_group_by_variables=[],selection_filters=[],interval=None):
    group_by_variables = input_group_by_variables
    group_df_variables = []
    for input_group_by_variable in input_group_by_variables:
      group_df_variables = group_df_variables + [input_group_by_variable]
    the_excluded_players=excluded_players()
    filters = selection_filters+the_excluded_players
    if interval == None:
        raw = (events().minimum(
        table,
        group_by=group_by_variables,
        target_property=target_variable,
        filters=filters,
        timeframe={'start': period_start.strftime('%Y-%m-%dT%H:%M:%S.%fZ'), 'end': period_finish.strftime('%Y-%m-%dT%H:%M:%S.%fZ') },
        timezone='utc'
        ))
        if len(group_df_variables) == 0:
            raw_df = pd.DataFrame([{'start': period_start, 'result': raw}])
        else:
            raw_df=pd.DataFrame(raw)
        raw_df=raw_df.rename(columns={'result' : result_variable_name})
        return raw_df
    else:
        raw = (events().minimum(
        table,
        target_property=target_variable,
        group_by=group_by_variables,
        filters=filters,
        timeframe={'start': period_start.strftime('%Y-%m-%dT%H:%M:%S.%fZ'), 'end': period_finish.strftime('%Y-%m-%dT%H:%M:%S.%fZ') },
        interval = interval,
        timezone='utc'))
        raw_df = pd.DataFrame(extract_timeinterval(raw))
        raw_df=raw_df.rename(columns={'result' : result_variable_name})

        if len(group_df_variables) == 0:
            return raw_df
        else:
          data_df = raw_df.groupby(group_df_variables+['start']).sum()
          return data_df