kavya006 kavya006

## tf_ts_classification_model.py
class Baseline(tf.keras.Model):
  def __init__(self, label_index=None):
    super().__init__()
    self.label_index = label_index

  def call(self, inputs):
    if self.label_index is None:
      return inputs
    result = inputs[:, :, self.label_index]
    # Convert the labels to one hot for the loss functions to be able to calculate the loss correctly.

## tf_ts_classification_plotting.py
def plot(self, model=None, plot_col=REGRESSORS[0], label_col=LABELS[0], max_subplots=3):
  inputs, labels = self.example
  plt.figure(figsize=(12, 8))

  # Earlier this code was inside the for loop. This code does not change based on the iteration. So I moved it outside the loop.
  plot_col_index = self.column_indices[plot_col]
  max_n = min(max_subplots, len(inputs))
  if self.label_columns:
    label_col_index = self.label_columns_indices.get(label_col, None)
  else:

## tf_ts_classification_data_preprocessing_update_datasets.py

def update_datasets(self:MultiSeriesWindowsGenerator, train_df:pd.DataFrame, val_df:pd.DataFrame, test_df:pd.DataFrame, norm:bool=False):
    # Store the raw data. We now get the features and targets separately from preprocess_dataset.
    # We will need to concat them together later to be backward compatible.

    train_df, train_targets = self.preprocess_dataset(train_df)
    val_df, val_targets = self.preprocess_dataset(val_df)
    test_df, test_targets = self.preprocess_dataset(test_df)

    # We should not normalize the targets. This is the reason why we had to separate the features and targets in the previous step.

## tf_ts_classification_data_preprocessing_preprocess_dataset.py
def preprocess_dataset(self:MultiSeriesWindowsGenerator, data:pd.DataFrame):
    try:
      if np.vstack(data.index).shape[1] != 1:
        data = data.reset_index()

      by = self.GROUPBY + [DATE]
      labels = self.regressor_columns + self.static_columns
      data = data.set_index(by).unstack(-1)
      features = tf.stack([data[label] for label in labels], axis=-1)
      targets = tf.stack([data[label] for label in self.label_columns], axis=-1)

## tf_ts_classification_data_generation.py
series = pd.concat([series_1, series_2, series_3], axis=0).reset_index(drop=True)
for label in LABELS:
  series[label] = pd.qcut(series[label], 4, labels=[0, 1, 2, 3]).astype(float)

## t2_complete.py
def print_t2(x="t2"):
    print(f"{x} called")

def print_t5(x="t5"):
    print(f"{x} called 594859..")

## pyproject.toml
[build-system]
requires = ["setuptools"]
build-backend = "setuptools.build_meta"

[metadata]
name = "project"
version = "0.0.1"

[tool.setuptools.packages]
find = {}  # Scan the project directory with the default parameters

## pyproject_3.toml
[tool.setuptools.packages]
find = {}  # Scan the project directory with the default parameters

## pyproject_2.toml

[metadata]
name = "project"
version = "0.0.1"

[project]
name = "project"
version = "0.0.1"
authors = [
  { name="your_name", email="your_email" },

## pyproject_1.toml
[build-system]
requires = ["setuptools"]
build-backend = "setuptools.build_meta"
	class Baseline(tf.keras.Model):
	def __init__(self, label_index=None):
	super().__init__()
	self.label_index = label_index

	def call(self, inputs):
	if self.label_index is None:
	return inputs
	result = inputs[:, :, self.label_index]
	# Convert the labels to one hot for the loss functions to be able to calculate the loss correctly.
	def plot(self, model=None, plot_col=REGRESSORS[0], label_col=LABELS[0], max_subplots=3):
	inputs, labels = self.example
	plt.figure(figsize=(12, 8))

	# Earlier this code was inside the for loop. This code does not change based on the iteration. So I moved it outside the loop.
	plot_col_index = self.column_indices[plot_col]
	max_n = min(max_subplots, len(inputs))
	if self.label_columns:
	label_col_index = self.label_columns_indices.get(label_col, None)
	else:

	def update_datasets(self:MultiSeriesWindowsGenerator, train_df:pd.DataFrame, val_df:pd.DataFrame, test_df:pd.DataFrame, norm:bool=False):
	# Store the raw data. We now get the features and targets separately from preprocess_dataset.
	# We will need to concat them together later to be backward compatible.

	train_df, train_targets = self.preprocess_dataset(train_df)
	val_df, val_targets = self.preprocess_dataset(val_df)
	test_df, test_targets = self.preprocess_dataset(test_df)

	# We should not normalize the targets. This is the reason why we had to separate the features and targets in the previous step.
	def preprocess_dataset(self:MultiSeriesWindowsGenerator, data:pd.DataFrame):
	try:
	if np.vstack(data.index).shape[1] != 1:
	data = data.reset_index()

	by = self.GROUPBY + [DATE]
	labels = self.regressor_columns + self.static_columns
	data = data.set_index(by).unstack(-1)
	features = tf.stack([data[label] for label in labels], axis=-1)
	targets = tf.stack([data[label] for label in self.label_columns], axis=-1)
	series = pd.concat([series_1, series_2, series_3], axis=0).reset_index(drop=True)
	for label in LABELS:
	series[label] = pd.qcut(series[label], 4, labels=[0, 1, 2, 3]).astype(float)
	def print_t2(x="t2"):
	print(f"{x} called")

	def print_t5(x="t5"):
	print(f"{x} called 594859..")
	[build-system]
	requires = ["setuptools"]
	build-backend = "setuptools.build_meta"

	[metadata]
	name = "project"
	version = "0.0.1"

	[tool.setuptools.packages]
	find = {} # Scan the project directory with the default parameters