Trying to run a Bayesian test

User Discussion
1

I’m trying to run a Bayesian test for an excel dataframe. When it runs, I get the following warning:

ABNORMAL_TERMINATION_IN_LNSRCH

 Line search cannot locate an adequate point after MAXLS
  function and gradient evaluations.
  Previous x, f and g restored.
 Possible causes: 1 error in function or gradient evaluation;
                  2 rounding error dominate computation.

The test returns a value so close to zero it can be despicable. In the dataframe the average increases about 7% when comparing the pre_period with the post_period. This result from the Bayesian test doesn’t make any sense.

## Teste Bayesiano para verificar a influência da pandemia no mercado Imobiliário

from causalimpact import CausalImpact

import pandas as pd

import numpy as np

from statsmodels.tsa.arima_process import arma_generate_sample

import matplotlib.pyplot as plt

import matplotlib

df = pd.read_excel("Dados_FIPE_Isolados_2018-2023.xlsx", sheet_name="Planilha1", header=0, index_col=0, parse_dates=True);

# Verificando se o DataFrame está vazio

if df.empty:

print("Erro: DataFrame está vazio!")

else:

print("DataFrame carregado com sucesso!")

print(df.head())

# Arredondando os valores para 2 casas decimais;

df = df.round (2);

df.index = pd.to_datetime(df.index) # Garantir que o índice é datetime

df.index = df.index.round('S') # Arredondar para o segundo mais próximo

#verificando o índice

print("index verificado com sucesso!")

#verificando o df

print(df.head(10));

#Definindo o período de pré e pós-intervenção com base no número das linhas

pre_period = ['2018-01-01', '2020-03-01']

post_period = ['2020-04-01', '2023-12-01']

#Rodando o teste

ts_impact = CausalImpact(df, pre_period, post_period)

ts_impact.run()

ts_impact.summary()

ts_impact.plot()

print(ts_impact.summary())

 warnings.warn(msg, FutureWarning)
                              Average        Cumulative
Actual                           6328            284803
Predicted                        6328            284766
95% CI                   [6327, 6329]  [284721, 284811]

Absolute Effect                     0                36
95% CI                         [1, 0]          [81, -8]

Relative Effect                  0.0%              0.0%
95% CI                  [0.0%, -0.0%]     [0.0%, -0.0%]

P-value                          0.0%
Prob. of Causal Effect         100.0%
                              Average        Cumulative
Actual                           6328            284803
Predicted                        6328            284766
95% CI                   [6327, 6329]  [284721, 284811]

Absolute Effect                     0                36
95% CI                         [1, 0]          [81, -8]

Relative Effect                  0.0%              0.0%
95% CI                  [0.0%, -0.0%]     [0.0%, -0.0%]

P-value                          0.0%
Prob. of Causal Effect         100.0%
None