def rocket_launch_count():
""" rocket count """
rocket_info = launch_data[['rocket']].to_dict(orient='index')
rocket_types = []
for _, value in rocket_info.items():
rocket_types.append(value['rocket']['rocket_name'])
rocket_type_series = pandas.Series(rocket_types)
rocket_type_counts = rocket_type_series.value_counts()
return rocket_type_counts
def successful_rocket_launch_count():
"""succesful rocket count """
successful_launches_df = launch_data[launch_data['launch_success'] == True]
successful_rocket_info = successful_launches_df[['rocket'
]].to_dict(orient='index')
successful_rocket_types = []
for _, value in successful_rocket_info.items():
successful_rocket_types.append(value['rocket']['rocket_name'])
successful_rocket_type_series = pandas.Series(successful_rocket_types)
successful_rocket_type_counts = successful_rocket_type_series.value_counts(
)
return successful_rocket_type_counts
def rocket_launch_stats():
""" average successful launches by rocket type"""
rocket_launch_averages = {}
for rocket_name, rocket_count, successful_rocket_count in zip(
rocket_launch_count().index.to_list(), rocket_launch_count(),
successful_rocket_launch_count()):
success_average = round((successful_rocket_count / rocket_count * 100),
2)
rocket_launch_averages[rocket_name] = success_average
rocket_launch_averages_series = pandas.Series(rocket_launch_averages)
rocket_launch_averages_plot = rocket_launch_averages_series.plot(
kind='barh').get_figure()
rocket_launch_stats = configure_graph(
series_plot=rocket_launch_averages_plot,
title='Successful Launch Averages by Rocket Name',
x_label='Averages (%)',
y_label='Name',
count_values=list(rocket_launch_averages_series))
return rocket_launch_stats
def site_usage():
""" return flights by rocket graph """
launch_site_info = launch_data[['launch_site']].to_dict(orient='index')
launch_site_locations = []
for _, value in launch_site_info.items():
launch_site_locations.append(value['launch_site']['site_name'])
launch_site_locations_series = pandas.Series(launch_site_locations)
launch_site_location_counts = launch_site_locations_series.value_counts()
launch_site_location_count_plot = launch_site_location_counts.plot(
kind='barh').get_figure()
launch_site_location_count_list = list(launch_site_location_counts)
launch_site_location_list_stats = configure_graph(
series_plot=launch_site_location_count_plot,
title='Number of times Launch Site were used',
x_label='Times Used',
y_label='Site',
count_values=launch_site_location_count_list)
return launch_site_location_list_stats
def curva_promedio(ansem,up,variedad,edad):
try:
valor=promedio[ansem,up,variedad,edad]['coeficiente']
return valor
except:
return 0
datos.ano_semana=datos.ano_semana.astype(str)
datos=datos.merge(semanas,on='ano_semana',how='left')
datos['mes_dato']=datos.dia.dt.month
recons=datos.sort_values(['concatenado','edad']).reset_index(drop=True)
lag_prod=10
for i in tqdm(range(1,lag_prod+1)):
strprod=str(i)+'sem_atras'
strconc=str(i)+'concat_atras'
recons[strprod]=recons['coeficiente'].shift(i)
recons[strconc]=recons['concatenado'].shift(i)
vald=str(i)+'valido'
recons[vald]=recons.apply(lambda ff: 1 if ff[strconc]==ff['concatenado'] else 0,axis=1)
recons[strprod]=recons.apply(lambda x: x[strprod] if x[vald]==1 else 0,axis=1)
recons.drop(columns={strconc},inplace=True)
recons.drop(columns={vald},inplace=True)
recons.drop(columns={'Unnamed: 0'},inplace=True)
recons=recons.merge(colores[['variedad','Color']],how='left',on='variedad')
#agrega la columna de curva estandar para cada variedad-finca
recons.ano_semana=(recons.ano_semana).astype(str)
recons['curva_metodo_finca'] = recons.apply(lambda x: curva_promedio(x['ano_semana'],x['finca'],x['variedad'],x['edad']),axis=1)
recons=recons[recons['tipo'].isin(['Minicarnation','Carnation'])]
recons.Color.fillna('NoColor',inplace=True)
recons['edad^2']=recons['edad']**2
recons['edad^3']=recons['edad']**3
#Red Neuronal
consolidado_rn=pd.DataFrame()
from sklearn.preprocessing import StandardScaler
from tensorflow import keras
from tensorflow.keras import layers
recons=recons[recons['dia']>='01/01/2018']
y_hat_rn=pd.Series(name='y_hat_falso')
for i in recons.tipo.unique():
for j in recons_test[recons_test['tipo']==i]['Color'].unique():
temp_test=recons_test[(recons_test['tipo']==i)&(recons_test['Color']==j)]
df_clean_test=pd.concat([temp_test[['edad','edad^2','edad^3','mes_dato','5sem_atras',
'6sem_atras','7sem_atras','8sem_atras','9sem_atras','10sem_atras',
#'11sem_atras','12sem_atras','13sem_atras','14sem_atras','15sem_atras',
'curva_metodo_finca','coeficiente']], pd.get_dummies(temp_test['variedad']), pd.get_dummies(temp_test['finca'])], axis=1)
df_clean_test.fillna(value=0,inplace=True)
y_real_test = df_clean_test.coeficiente
X_real_test = df_clean_test.drop('coeficiente', axis=1)
temp=recons[(recons['tipo']==i)&(recons['Color']==j)]
temp=temp[temp['variedad'].isin(temp_test['variedad'].unique())]
temp=temp[temp['finca'].isin(temp_test['finca'].unique())]
df_clean=pd.concat([temp[['edad','edad^2','edad^3','mes_dato','5sem_atras',
'6sem_atras','7sem_atras','8sem_atras','9sem_atras','10sem_atras',
#'11sem_atras','12sem_atras','13sem_atras','14sem_atras','15sem_atras',
'curva_metodo_finca','coeficiente']], pd.get_dummies(temp['variedad']), pd.get_dummies(temp['finca'])], axis=1)
df_clean.fillna(value=0,inplace=True)
y = df_clean.coeficiente
X = df_clean.drop('coeficiente', axis=1)
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.33, random_state=42)
scaler = StandardScaler()
X_train_std = pd.DataFrame(scaler.fit_transform(X_train), columns = X_train.columns)
neurons = 256
model = keras.Sequential([layers.Dense(neurons, activation='relu', input_shape=[len(X_train_std.columns)]),
layers.Dense(neurons,activation='relu'),
layers.Dense(1,activation='relu')]) #Capa salida
model.compile(loss='mse', optimizer = 'adam')
history = model.fit(X_train_std, y_train, epochs=100, validation_split = 0.2, verbose=0,batch_size=100)
X_norm = scaler.transform(X_real_test)
indice=X_real_test.reset_index()['index']
y_hat=model.predict(X_norm)
y_hat=pd.Series(y_hat[0:,0],name='y_hat')
y_hat.index=X_real_test.index
y_hat_rn=pd.concat([y_hat_rn,y_hat],axis=1)
y_hat_rn.drop(columns={'y_hat_falso'},inplace=True)
ser_y_hat=np.sum(y_hat_rn,axis=1)
y_hat_rn['y_hat_red_n']=ser_y_hat
validacion_y_hat=y_hat_rn[['y_hat_red_n']]
validacion_final=pd.concat([recons_test,validacion_y_hat],axis=1)
def main():
SetupLogger()
logging.debug("now is %s", datetime.datetime.now())
logging.getLogger().setLevel(logging.INFO)
cmdLineParser = argparse.ArgumentParser("api tests")
cmdLineParser.add_argument("-a",
"--account",
action="store",
type=str,
dest="account",
help="The account number to use")
cmdLineParser.add_argument("-p",
"--port",
action="store",
type=int,
dest="port",
default=7497,
help="The TCP port to use")
cmdLineParser.add_argument("-C",
"--global-cancel",
action="store_true",
dest="global_cancel",
default=False,
help="whether to trigger a globalCancel req")
args = cmdLineParser.parse_args()
print("Using args", args)
logging.debug("Using args %s", args)
# <<<<<<<< BEGIN
# Be sure to read requirements:
# https://interactivebrokers.github.io/tws-api/introduction.html#requirements
# AND IMPORTANT TWS Setup
# https://interactivebrokers.github.io/tws-api/initial_setup.html
try:
bot1 = sincBot()
bot1.clientId = random.randint(
1, 32) #This can be random or 0 for master client
bot1.account = args.account #Set this to your account number
bot1.portNum = args.port #Set this to the portnumber you configured in Trader
#
#Read constraints in API Doc:
# https://interactivebrokers.github.io/tws-api/historical_bars.html
bot1.duration = '1 M' #How far back to go
bot1.barSize = '1 day'
bot1.type = 'TRADES'
bot1.timeStr = datetime.datetime.now().strftime(
'%Y%m%d %H:%M:%S') #From today
bot1.reconnect()
# Add interesting symbols here
symbols = numpy.array(['AAPL', 'MSFT'])
data = pandas.Series(symbols)
bot1.start(data) #This begine executions
print(bot1.cache)
bot1.stop()
sys.exit(1)
# ! [clientrun]
except:
raise
finally:
bot1.dumpTestCoverageSituation()
bot1.dumpReqAnsErrSituation()