def plot_sample_composition_most_abundant_taxa(
self,
mean_taxa: float = None,
taxa_number: int = 20,
taxa_level: str = "species",
cluster_samples: bool = True,
samples_order: List[str] = None,
**kwargs,
):
"""
Plot taxa composition of samples for most abundant taxa.
Args:
mean_taxa: mean threshold to be kept for analysis
taxa_number: number of taxa to plot
taxa_level: Taxonomy level
cluster_samples: use clustering (skipped by samples_order)
samples_order: list of samples to force ordering for visualization
"""
data_df = self.df
if mean_taxa is not None:
data_df = TaxonomyMeanFiltering(data_df, mean_taxa).filtered_df
data_df = self._compute_abundances_for_n_taxa(data_df, taxa_number, taxa_level)
if samples_order is not None:
data_df = data_df.loc[:, samples_order]
elif cluster_samples:
data_df = self._cluster_samples(data_df)
# Make graph
graph = MatrixBarGraph(data_df)
# Plotting options
default_plotting_options = {
"layout": {
"title": f"{taxa_level.capitalize()} composition for the top {data_df.drop('Others').shape[0]} most abundant species across samples", # noqa
"xaxis_title": "Samples",
"yaxis_title": "Percentage",
"legend": {"traceorder": "normal"},
}
}
if mean_taxa is not None:
default_plotting_options["layout"][
"title"
] = "{} (mean among samples > {})".format(
default_plotting_options["layout"]["title"], mean_taxa
)
plotting_options = merge_dict(
kwargs.pop("plotting_options", {}), default_plotting_options
)
graph.plot_one_graph(plotting_options=plotting_options, **kwargs)
def plot_most_abundant_taxa(
self,
mean_taxa: float = None,
taxa_number: int = 20,
taxa_level: str = "species",
**kwargs,
):
"""
Plot bar chart of most abundant taxa (total sum of abundance).
The plot represents percentage of sample with the corresponding taxa
ordered from most abundant to less abundant.
Args:
mean_taxa: Mean threshold to be kept for analysis
taxa_number: Number of taxa to plot
taxa_level: Taxonomy level
"""
data_df = self.df
if mean_taxa is not None:
data_df = TaxonomyMeanFiltering(data_df, mean_taxa).filtered_df
percentage_presence = self._get_percentage_presence(
data_df, taxa_level, taxa_number
)
taxa_number = len(percentage_presence)
# Make graph
graph = BarGraph(percentage_presence.iloc[::-1])
# Plotting options
default_plotting_options = {
"layout": {
"title": f"{taxa_number} most abundant {taxa_level} - Total sum of abundances",
"xaxis_title": "Percentage Sample",
"yaxis_title": taxa_level.capitalize(),
}
}
if mean_taxa is not None:
default_plotting_options["layout"][
"title"
] = "{} (mean among samples > {})".format(
default_plotting_options["layout"]["title"], mean_taxa
)
plotting_options = merge_dict(
kwargs.pop("plotting_options", {}), default_plotting_options
)
graph.plot_one_graph(
orientation="h", plotting_options=plotting_options, **kwargs
)
def test_simple_dicts(self):
first_dict = {"a": 3, "b": 4}
second_dict = {"a": 5, "c": 7}
expected_dict = {"a": 3, "b": 4, "c": 7}
self.assertDictEqual(merge_dict(first_dict, second_dict),
expected_dict)
def test_with_multiple_subdict(self):
first_dict = {"a": 3, "sub": {"b": 4, "subsub": {"z": 3}}}
second_dict = {"a": 4, "sub": {"b": 7, "d": 6, "subsub": {"z": 5}}}
expected_dict = {"a": 3, "sub": {"b": 4, "d": 6, "subsub": {"z": 3}}}
self.assertDictEqual(merge_dict(first_dict, second_dict),
expected_dict)
def test_with_subdict_variant_two(self):
first_dict = {"a": 3}
second_dict = {"a": 4, "sub": {"b": 4, "c": 6}}
expected_dict = {"a": 3, "sub": {"b": 4, "c": 6}}
self.assertDictEqual(merge_dict(first_dict, second_dict),
expected_dict)