def find_all_process_most_frequent(varas_group):
macro_trace_processes = {}
for vara in varas_group:
df_vara = df_log[df_log['case: orgao'] == vara]
p = PreProcess(df=df_vara)
p.select_desired_columns()
p.filter_outlier_timestamp()
p.filter_outlier_movements(lower=0.05, upper=0.95)
p.filter_outlier_trace_time(lower=0.05, upper=0.95)
l = Log(df_log=p.df_log.sort_values('time:timestamp'))
all_macro_trace = find_all_macro_trace(l.log, macrosteps)
for tran in all_macro_trace:
if tran not in macro_trace_processes:
macro_trace_processes[tran] = 0
macro_trace_processes[tran] += 1
macro_trace_processes = {k: v for k, v in \
sorted(macro_trace_processes.items(),
key=lambda item: item[1])}
macro_trace_processes
from log.Log import Log
from log.PreProcess import PreProcess
from process.MacroSteps import MacroSteps
file_path = '/home/vercosa/Documentos/bases_desafio_cnj/'+\
'versao6/resultado_1_0_null.csv'
# file_path = '/home/vercosa/Documentos/bases_desafio_cnj/'+\
# 'versao5/version_5.csv'
movement_path = '/home/vercosa/Insync/doutorado/hackaton_cnj/' + \
'projeto_git/desafio_cnj/data/interim/df_movimentos.csv'
pp = PreProcess(file_location=file_path)
# pp.select_desired_columns()
pp.filter_outlier_timestamp()
pp.map_movements(movement_path)
# filter first movement to "Distribuição"
df_log = pp.df_log
df_log
df_first = df_log.groupby('case:concept:name', as_index=False).\
agg({'time:timestamp':'min', 'concept:name':'first'})
# df_first.groupby('concept:name', as_index=False).count().\
# sort_values(by='case:concept:name', ascending=False)
from pm4py.algo.discovery.dfg import algorithm as dfg_discovery
from pm4py.statistics.traces.log import case_statistics
from pm4py.algo.discovery.dfg import parameters
from pm4py.visualization.dfg import visualizer as dfg_visualization
from log.Log import Log
from log.Log import INTENSE_FILTERING
from discovery.DFG import DFG
import visualization.Visualizer as Visualizer
from log.PreProcess import PreProcess
file_path = '/home/vercosa/Documentos/bases_desafio_cnj/'+\
'log_vara_2.csv'
p = PreProcess(file_location=file_path)
p.select_desired_columns()
p.filter_outlier_timestamp()
p.filter_outlier_movements(lower=0.01, upper=0.99)
p.filter_outlier_trace_time(lower=0.01, upper=0.99)
l = Log(df_log=p.df_log.sort_values('time:timestamp'))
# l.filter_variants(1.1)
dfg = DFG(l.log,
parameters={parameters.Parameters.AGGREGATION_MEASURE:'mean'},
variant=dfg_discovery.Variants.FREQUENCY)
dfg.filter_activities(number_act=10)
dfg.filter_edges(percentage=0.3)
engine='python')
df_group = df_group[df_group['process_count'] > 50]
varas_group = df_group['case: orgao'].tolist()
dfg_dict = {}
act_dict = {}
act_dict_temp = {}
similar_group = {}
similar_group_ref = {}
for vara in varas_group:
df_temp = df_log[df_log['case: orgao'] == vara]
p = PreProcess(df=df_temp)
p.select_desired_columns()
p.filter_outlier_timestamp()
p.filter_outlier_movements(lower=0.01, upper=0.99)
p.filter_outlier_trace_time(lower=0.01, upper=0.99)
l = Log(df_log=p.df_log.sort_values('time:timestamp'))
# l.filter_variants(1)
dfg = DFG(l.log)
dfg.filter_activities(number_act=10)
dfg.filter_edges(percentage=0.3)
dfg_dict[vara] = list(dfg.dfg)
for vara in dfg_dict.keys():
act_dict[vara] = (list(set([item for sublist in dfg_dict[vara] \
for item in sublist])),
'projeto_git/desafio_cnj/data/interim/df_movimentos.csv'
vara = '2ª VARA C�?VEL DE SANTANA'
macrosteps = [
'Distribuição',
'Conclusão',
'Despacho',
'Decisão',
'Julgamento',
'Trânsito em julgado',
'Baixa/Arquivamento',
]
pp = PreProcess(file_location=file_path)
pp.select_desired_columns()
pp.filter_outlier_timestamp()
pp.map_movements(movement_path)
df_log = pp.df_log
df_vara = df_log[df_log['case: orgao'] == vara]
pp_vara = PreProcess(df=df_vara)
pp_vara.filter_outlier_movements(lower=0.05, upper=0.95)
pp_vara.filter_outlier_trace_time(lower=0.05, upper=0.95)
log = Log(df_log=pp_vara.df_log.sort_values('time:timestamp'))
median_case_duration = case_statistics.\
get_median_caseduration(log.log, parameters={