vaep.analyzers package#
General classes formalizing an experiment.
- class vaep.analyzers.Analysis[source]#
Bases:
SimpleNamespace
Submodules#
vaep.analyzers.analyzers module#
A collection of Analyzers to perform certain type of analysis.
- class vaep.analyzers.analyzers.AnalyzePeptides(data: DataFrame, is_log_transformed: bool = False, is_wide_format: bool = True, ind_unstack: str = '')[source]#
Bases:
SimpleNamespaceNamespace for current analysis
- df#
current eagerly loaded data in wide format only: sample index, features in columns
- Type:
- stats#
Some statistics of certain aspects. Normally each will be a DataFrame.
- Type:
- Many more attributes are set dynamically depending on the concrete analysis.
- property df_long#
- property df_wide#
- property fname_stub#
- classmethod from_csv(fname: str, nrows: Optional[int] = None, index_col: Union[int, str, List] = 'Sample ID', verify_fname: bool = False, usecols=None, **kwargs)[source]#
- classmethod from_pickle(fname: str, index_col: Union[int, str, List] = 'Sample ID', verify_fname: bool = False, usecols=None, **kwargs)[source]#
- get_consecutive_dates(n_samples, seed=42)[source]#
Select n consecutive samples using a seed.
Updated the original DataFrame attribute: df
- get_dectection_limit()[source]#
Compute information on detection limit in dataset.
- Returns:
Information on detection limit
- Return type:
- log_transform(log_fct: ufunc)[source]#
Log transform data in-place.
- Parameters:
log_fct (np.ufunc) – Numpy log-function
- Raises:
Exception – if data has been previously log-transformed.
- plot_pca()[source]#
Create principal component plot with three heatmaps showing instrument, degree of non NA data and sample by date.
- to_long_format(colname_values: str = 'intensity', index_name: str = 'Sample ID', inplace: str = False) DataFrame[source]#
[summary]
- Parameters:
colname_values (str, optional) – New column name for values in matrix, by default ‘intensity’
index_name (str, optional) – Name of column to assign as index (based on long-data format), by default ‘Sample ID’
inplace (bool, optional) – Assign result to df_long (False), or to df (True) attribute, by default False
- Returns:
Data in long-format as DataFrame
- Return type:
pd.DataFrame
- class vaep.analyzers.analyzers.LatentAnalysis(latent_space: DataFrame, meta_data: DataFrame, model_name: str, fig_size: Tuple[int, int] = (15, 15), folder: Optional[Path] = None)[source]#
Bases:
Analysis
- vaep.analyzers.analyzers.cast_object_to_category(df: DataFrame) DataFrame[source]#
Cast object columns to category dtype.
- Parameters:
df (pd.DataFrame) – DataFrame with columns
- Returns:
DataFrame with category columns instead of object columns.
- Return type:
pd.DataFrame
- vaep.analyzers.analyzers.corr_lower_triangle(df)[source]#
Compute the correlation matrix, returning only unique values.
- vaep.analyzers.analyzers.plot_date_map(df, ax, dates: Optional[Series] = None, meta: Optional[Series] = None, title: str = 'by date', fontsize=8, size=2)[source]#
- vaep.analyzers.analyzers.plot_scatter(df, ax, meta: Series, feat_name_display: str = 'features', title: Optional[str] = None, alpha=0.5, fontsize=8, size=2)[source]#
- vaep.analyzers.analyzers.scatter_plot_w_dates(ax, df, dates=None, marker=None, errors='raise', size=2)[source]#
plot first vs. second column in DataFrame. Use dates to color data.
- errors{‘ignore’, ‘raise’, ‘coerce’}, default ‘raise’
Passed on to pandas.to_datetime - If ‘raise’, then invalid parsing will raise an exception. - If ‘coerce’, then invalid parsing will be set as NaT. - If ‘ignore’, then invalid parsing will return the input.
vaep.analyzers.compare_predictions module#
- vaep.analyzers.compare_predictions.load_predictions(pred_files: List, shared_columns=['observed'])[source]#
- vaep.analyzers.compare_predictions.load_single_csv_pred_file(fname: str | Path, value_name: str = 'intensity') pd.Series[source]#
- Load a single pred file from a single model.
Last column are measurments, other are index.
- vaep.analyzers.compare_predictions.load_split_prediction_by_modelkey(experiment_folder: Path, split: str, model_keys: list[str], allow_missing=False, shared_columns: list[str] = None)[source]#
Load predictions from a list of models.
- Parameters:
experiment_folder (Path) – Path to experiment folder
split (str) – which split of simulated data to load
model_keys (List) – List of model keys to be loaded
allow_missing (bool, optional) – Ignore missing pred files of requested model, default False
shared_columns (List, optional) – List of columns that are shared between all models, by default None
- Returns:
Prediction data frame with shared columns and model predictions
- Return type:
pd.DataFrame
vaep.analyzers.diff_analysis module#
- class vaep.analyzers.diff_analysis.Cutoffs(feat_completness_over_samples, min_feat_in_sample)#
Bases:
tuple- feat_completness_over_samples#
Alias for field number 0
- min_feat_in_sample#
Alias for field number 1