STY: Correct flake8 issues

Author: bashtage

INSTALL.txt

@@ -10,7 +10,7 @@ numpy >= 1.18
     www.numpy.org
 
 scipy >= 1.4
-    
+
     www.scipy.org
 
 pandas >= 1.0
@@ -85,7 +85,7 @@ On Ubuntu you can get dependencies through:
     python -m pip install cython pandas
 
 Alternatively, you can install from the NeuroDebian repository:
-    
+
     http://neuro.debian.net
 
 

statsmodels/datasets/star98/r_glm.s

@@ -21,7 +21,7 @@ m1 <-  glm(cbind(PR50M,MATHTOT-PR50M) ~ LOWINC + PERASIAN + PERBLACK + PERHISP +
 #as.numeric(m1$coef)
 #as.numeric(sqrt(diag(vcov(m1))))
 results <- summary.glm(m1)
-    
+
 #star.logit.fit3 <- glm(cbind(PR50M,MATHTOT-PR50M) ~ LOWINC + PERASIAN + PERBLACK + PERHISP +
 #    PERMINTE + AVYRSEXP + AVSAL + PERSPEN + PTRATIO + PCTAF + PCTCHRT + PCTYRRND +
 #    PERMINTE.AVYRSEXP + PERMINTE.AVSAL + AVYRSEXP.AVSAL + PERSPEN.PTRATIO + PERSPEN.PCTAF + 

statsmodels/tsa/ar_model.py

@@ -164,7 +164,7 @@ class AutoReg(tsa_model.TimeSeriesModel):
     -----
     See the notebook `Autoregressions
     <../examples/notebooks/generated/autoregressions.html>`__ for an overview.
-    
+
     Examples
     --------
     >>> import statsmodels.api as sm

statsmodels/tsa/filters/bk_filter.py

@@ -55,7 +55,7 @@ def bkfilter(x, low=6, high=32, K=12):
 
     See the notebook `Time Series Filters
     <../examples/notebooks/generated/tsa_filters.html>`__ for an overview.
-    
+
     References
     ----------
     Baxter, M. and R. G. King. "Measuring Business Cycles: Approximate

statsmodels/tsa/filters/cf_filter.py

@@ -55,7 +55,7 @@ def cffilter(x, low=6, high=32, drift=True):
     -----
     See the notebook `Time Series Filters
     <../examples/notebooks/generated/tsa_filters.html>`__ for an overview.
-    
+
     Examples
     --------
     >>> import statsmodels.api as sm

statsmodels/tsa/filters/hp_filter.py

@@ -58,7 +58,7 @@ def hpfilter(x, lamb=1600):
 
     See the notebook `Time Series Filters
     <../examples/notebooks/generated/tsa_filters.html>`__ for an overview.
-    
+
     References
     ----------
     Hodrick, R.J, and E. C. Prescott. 1980. "Postwar U.S. Business Cycles: An

statsmodels/tsa/forecasting/stl.py

@@ -118,8 +118,9 @@ class STLForecast:
       both in- and out-of-sample predictions.
 
     See the notebook `Seasonal Decomposition
-    <../examples/notebooks/generated/stl_decomposition.html>`__ for an overview.
-    
+    <../examples/notebooks/generated/stl_decomposition.html>`__ for an
+    overview.
+
     Examples
     --------
     >>> import numpy as np

statsmodels/tsa/holtwinters/_exponential_smoothers.pyx

@@ -238,7 +238,7 @@ cdef object holt_win_init(double[::1] x, HoltWintersArgs hw_args):
     cdef Py_ssize_t i, n, idx = 0
     cdef double[::1] p, s
     cdef long[::1] xi
-    
+
     p = hw_args._p
     xi = hw_args._xi
     n = p.shape[0]

statsmodels/tsa/holtwinters/model.py

@@ -197,7 +197,7 @@ class ExponentialSmoothing(TimeSeriesModel):
     See the notebook `Exponential Smoothing
     <../examples/notebooks/generated/exponential_smoothing.html>`__
     for an overview.
-    
+
     References
     ----------
     .. [1] Hyndman, Rob J., and George Athanasopoulos. Forecasting: principles
@@ -1532,7 +1532,7 @@ class SimpleExpSmoothing(ExponentialSmoothing):
     See the notebook `Exponential Smoothing
     <../examples/notebooks/generated/exponential_smoothing.html>`__
     for an overview.
-    
+
     References
     ----------
     .. [1] Hyndman, Rob J., and George Athanasopoulos. Forecasting: principles
@@ -1686,7 +1686,7 @@ class Holt(ExponentialSmoothing):
     See the notebook `Exponential Smoothing
     <../examples/notebooks/generated/exponential_smoothing.html>`__
     for an overview.
-    
+
     References
     ----------
     .. [1] Hyndman, Rob J., and George Athanasopoulos. Forecasting: principles

statsmodels/tsa/innovations/_arma_innovations.pyx.in

@@ -51,7 +51,7 @@ cdef {{prefix}}toeplitz(int n, int offset0, int offset1,
                         {{cython_type}} [:, :] out_matrix):
     """
     toeplitz(int n, int o0, int o1, {{cython_type}} [:] in_column, {{cython_type}} [:, :] out_matrix
-    
+
     Construct a Toeplitz block in a matrix in place
 
     Parameters
@@ -93,7 +93,7 @@ cpdef {{prefix}}arma_transformed_acovf_fast({{cython_type}} [:] ar,
                                             {{cython_type}} [:] arma_acovf):
     """
     arma_transformed_acovf_fast({{cython_type}} [:] ar, {{cython_type}} [:] ma, {{cython_type}} [:] arma_acovf)
-    
+
     Quickly construct the autocovariance matrix for a transformed process.
 
     Using the autocovariance function for an ARMA process, constructs the
@@ -201,7 +201,7 @@ cpdef {{prefix}}arma_innovations_algo_fast(int nobs,
                                            {{cython_type}} [:] acovf2):
     """
     arma_innovations_algo_fast(int nobs, {{cython_type}} [:] ar_params, {{cython_type}} [:] ma_params, {{cython_type}} [:, :] acovf, {{cython_type}} [:] acovf2)
-    
+
     Quickly apply innovations algorithm for an ARMA process.
 
     Parameters
@@ -304,7 +304,7 @@ cpdef {{prefix}}arma_innovations_filter({{cython_type}} [:] endog,
                                         {{cython_type}} [:, :] theta):
     """
     arma_innovations_filter({{cython_type}} [:] endog, {{cython_type}} [:] ar_params, {{cython_type}} [:] ma_params, {{cython_type}} [:, :] theta):
-    
+
     Innovations filter for an ARMA process.
 
     Parameters
@@ -375,7 +375,7 @@ cpdef {{prefix}}arma_innovations({{cython_type}} [:] endog,
                                       {{cython_type}} sigma2):
     """
     arma_innovations({{cython_type}} [:] endog, {{cython_type}} [:] ar_params, {{cython_type}} [:] ma_params):
-    
+
     Compute innovations and variances based on an ARMA process.
 
     Parameters

statsmodels/tsa/regime_switching/markov_autoregression.py

@@ -88,7 +88,7 @@ class MarkovAutoregression(markov_regression.MarkovRegression):
     See the notebook `Markov switching autoregression
     <../examples/notebooks/generated/markov_autoregression.html>`__
     for an overview.
-    
+
     References
     ----------
     Kim, Chang-Jin, and Charles R. Nelson. 1999.

statsmodels/tsa/regime_switching/markov_regression.py

@@ -72,8 +72,9 @@ class MarkovRegression(markov_switching.MarkovSwitching):
     ones.
 
     See the notebook `Markov switching dynamic regression
-    <../examples/notebooks/generated/markov_regression.html>`__ for an overview.
-    
+    <../examples/notebooks/generated/markov_regression.html>`__ for an
+    overview.
+
     References
     ----------
     Kim, Chang-Jin, and Charles R. Nelson. 1999.

statsmodels/tsa/statespace/_filters/_conventional.pyx.in

@@ -256,7 +256,7 @@ cdef int {{prefix}}updating_conventional({{prefix}}KalmanFilter kfilter, {{prefi
         {{cython_type}} alpha = 1.0
         {{cython_type}} beta = 0.0
         {{cython_type}} gamma = -1.0
-    
+
     # #### Filtered state for time t
     # $a_{t|t} = a_t + P_t Z_t' F_t^{-1} v_t$  
     # $a_{t|t} = 1.0 * \\#_1 \\#_2 + 1.0 a_t$

statsmodels/tsa/statespace/_filters/_univariate.pyx.in

@@ -565,7 +565,7 @@ cdef void {{prefix}}companion_predicted_state_cov({{prefix}}KalmanFilter kfilter
         {{cython_type}} tmp
 
     # $P_{t+1} = T_t P_{t,n} T_t' + Q_t^*$
-    
+
     # `tmp0` array used here, dimension $(p \times m)$  
     # $\\#_0 = \phi_t P_{t|t} $
 

statsmodels/tsa/statespace/_kalman_filter.pxd

@@ -196,7 +196,7 @@ cdef class sKalmanFilter(object):
     # ### Define some constants
     cdef readonly int k_endog, k_states, k_posdef, k_endog2, k_states2, k_posdef2, k_endogstates, k_statesposdef
     cdef readonly int ldwork
-    
+
     cdef allocate_arrays(self)
     cdef void set_dimensions(self)
     cpdef set_filter_method(self, int filter_method, int force_reset=*)
@@ -350,7 +350,7 @@ cdef class dKalmanFilter(object):
     # ### Define some constants
     cdef readonly int k_endog, k_states, k_posdef, k_endog2, k_states2, k_posdef2, k_endogstates, k_statesposdef
     cdef readonly int ldwork
-    
+
     cdef allocate_arrays(self)
     cdef void set_dimensions(self)
     cpdef set_filter_method(self, int filter_method, int force_reset=*)
@@ -505,7 +505,7 @@ cdef class cKalmanFilter(object):
     # ### Define some constants
     cdef readonly int k_endog, k_states, k_posdef, k_endog2, k_states2, k_posdef2, k_endogstates, k_statesposdef
     cdef readonly int ldwork
-    
+
     cdef allocate_arrays(self)
     cdef void set_dimensions(self)
     cpdef set_filter_method(self, int filter_method, int force_reset=*)
@@ -660,7 +660,7 @@ cdef class zKalmanFilter(object):
     # ### Define some constants
     cdef readonly int k_endog, k_states, k_posdef, k_endog2, k_states2, k_posdef2, k_endogstates, k_statesposdef
     cdef readonly int ldwork
-    
+
     cdef allocate_arrays(self)
     cdef void set_dimensions(self)
     cpdef set_filter_method(self, int filter_method, int force_reset=*)

statsmodels/tsa/statespace/_kalman_filter.pyx.in

@@ -366,7 +366,7 @@ cdef class {{prefix}}KalmanFilter(object):
     # ### Define some constants
     # cdef readonly int k_endog, k_states, k_posdef, k_endog2, k_states2, k_endogstates
     # cdef readonly ldwork
-    
+
     def __init__(self,
                  {{prefix}}Statespace model,
                  int filter_method=FILTER_CONVENTIONAL,

statsmodels/tsa/statespace/_kalman_smoother.pxd

@@ -468,7 +468,7 @@ cdef class zKalmanSmoother(object):
     )
 
     # cdef readonly int k_endog, k_states, k_posdef, k_endog2, k_states2, k_posdef2, k_endogstates, k_statesposdef
-    
+
     cdef allocate_arrays(self)
     cdef int check_filter_method_changed(self)
     cdef int reset_filter_method(self, int force_reset=*)

statsmodels/tsa/statespace/_kalman_smoother.pyx.in

@@ -237,7 +237,7 @@ cdef class {{prefix}}KalmanSmoother(object):
 
     # ### Define some constants
     # cdef readonly int k_endog, k_states, k_posdef, k_endog2, k_states2, k_posdef2, k_endogstates, k_statesposdef
-    
+
     def __init__(self,
                  {{prefix}}Statespace model,
                  {{prefix}}KalmanFilter kfilter,

statsmodels/tsa/statespace/_representation.pxd

@@ -13,7 +13,7 @@ cimport numpy as np
 cdef class sStatespace(object):
     # Statespace dimensions
     cdef readonly int nobs, k_endog, k_states, k_posdef
-    
+
     # Statespace representation matrices
     cdef readonly np.float32_t [::1,:] obs, obs_intercept, state_intercept
     cdef readonly np.float32_t [:] initial_state
@@ -90,7 +90,7 @@ cdef class sStatespace(object):
 cdef class dStatespace(object):
     # Statespace dimensions
     cdef readonly int nobs, k_endog, k_states, k_posdef
-    
+
     # Statespace representation matrices
     cdef readonly np.float64_t [::1,:] obs, obs_intercept, state_intercept
     cdef readonly np.float64_t [:] initial_state
@@ -167,7 +167,7 @@ cdef class dStatespace(object):
 cdef class cStatespace(object):
     # Statespace dimensions
     cdef readonly int nobs, k_endog, k_states, k_posdef
-    
+
     # Statespace representation matrices
     cdef readonly np.complex64_t [::1,:] obs, obs_intercept, state_intercept
     cdef readonly np.complex64_t [:] initial_state
@@ -244,7 +244,7 @@ cdef class cStatespace(object):
 cdef class zStatespace(object):
     # Statespace dimensions
     cdef readonly int nobs, k_endog, k_states, k_posdef
-    
+
     # Statespace representation matrices
     cdef readonly np.complex128_t [::1,:] obs, obs_intercept, state_intercept
     cdef readonly np.complex128_t [:] initial_state

statsmodels/tsa/statespace/_representation.pyx.in

@@ -81,7 +81,7 @@ cdef class {{prefix}}Statespace(object):
     # `k_posdef` $\equiv r$ is the dimension of the state shocks  
     # *Old notation: T, n, k, g*
     # cdef readonly int nobs, k_endog, k_states, k_posdef
-    
+
     # `obs` $\equiv y_t$ is the **observation vector** $(p \times T)$  
     # `design` $\equiv Z_t$ is the **design vector** $(p \times m \times T)$  
     # `obs_intercept` $\equiv d_t$ is the **observation intercept** $(p \times T)$  

statsmodels/tsa/statespace/_smoothers/_alternative.pyx.in

@@ -240,7 +240,7 @@ cdef int {{prefix}}smoothed_disturbances_alternative({{prefix}}KalmanSmoother sm
                       &alpha, smoother._tmp0, &kfilter.k_states,
                               &smoother.scaled_smoothed_estimator[0, smoother.t+1], &inc,
                       &beta, smoother._smoothed_state_disturbance, &inc)
-    
+
     if smoother.smoother_output & SMOOTHER_DISTURBANCE_COV:
         # $\\#_00 = K_t H_t$  
         # $(m \times p) = (m \times p) (p \times p)$  

statsmodels/tsa/statespace/_smoothers/_conventional.pyx.in

@@ -195,7 +195,7 @@ cdef int {{prefix}}smoothed_disturbances_missing_conventional({{prefix}}KalmanSm
     # TODO this is not explicitly addressed in Durbin and Koopman Chapter 4
     # or in Koopman (1993) - need to find a source for if this is correct
     # Note: this is what the MATLAB ssm toolbox does, also
-    
+
     # Smoothed measurement disturbances have unconditional expected
     # value of 0, so no need to do anything
 

statsmodels/tsa/statespace/tests/results/frbny_nowcast/Nowcasting/functions/dfm.m

@@ -1085,7 +1085,7 @@
 
 end
 
-    
+
 function [y,C,R,L]  = MissData(y,C,R)
 % Syntax:
 % Description:

statsmodels/tsa/statespace/tests/results/frbny_nowcast/Nowcasting/functions/summarize.m

@@ -20,10 +20,10 @@ function summarize(X,Time,Spec,vintage)
 fprintf([repmat('-',1,130) '\n']);
 
 for i = 1:N
-    
+
     % time indexes for which there are observed values for series i
     t_obs = ~isnan(X(:,i));
-    
+
     data_series = Spec.SeriesName{i};
     if length(data_series) > 30
         data_series = [data_series(1:27) '...'];
@@ -81,7 +81,7 @@ function summarize(X,Time,Spec,vintage)
             data_series,num_obs,units_transformed,frequency,mean_series,stdv_series,min_series,max_series);
     fprintf('%30s | %17s    %12s    %10s    %8s    %8s    %8s    %8s \n',...
             series_id,date_range,'','','','',min_date,max_date);
-    
+
 end
 
 fprintf('\n\n\n');