def _plot_and_save_single_group_worker(i, group, filename, cfg, d_fits):
set_style() # set style in child process
# try:
fig = plot_single_group(group, cfg, d_fits)
filename = filename_to_tmp_file(filename) + f"__{i:06d}.pdf"
utils.init_parent_folder(filename)
fig.savefig(filename)
return None
def save(self, df, filename, key, metadata=None):
utils.init_parent_folder(filename)
if metadata is None:
metadata = {}
with warnings.catch_warnings():
message = "object name is not a valid Python identifier"
warnings.filterwarnings("ignore", message=message)
with HDFStore(filename, mode="a") as hdf:
hdf.append(key, df, format="table", data_columns=True)
hdf.get_storer(key).attrs.metadata = metadata
def plot_fit_results(all_fit_results, cfg, N_alignments_mins=[-1]):
if not isinstance(N_alignments_mins, (list, tuple)):
N_alignments_mins = [N_alignments_mins]
if not 0 in N_alignments_mins:
N_alignments_mins = [0] + N_alignments_mins
filename = cfg.filename_plot_fit_results
if utils.is_pdf_valid(filename,
cfg.force_plots,
N_pages=len(N_alignments_mins)):
logger.info(f"Plot of fit results already exist.")
return None
set_style()
logger.info(f"Plotting fit results.")
n_sigma_lim, D_max_lim = find_fit_results_limits(all_fit_results)
utils.init_parent_folder(filename)
with PdfPages(filename) as pdf:
for N_alignments_min in N_alignments_mins:
# break
fig = plot_fit_results_single_N_aligment(
all_fit_results,
cfg,
N_alignments_min=N_alignments_min,
n_sigma_lim=n_sigma_lim,
D_max_lim=D_max_lim,
)
if fig:
pdf.savefig(fig, bbox_inches="tight", pad_inches=0.1)
# pdf.savefig(fig)
d = pdf.infodict()
d["Title"] = "Error Rate Distributions"
d["Author"] = "Christian Michelsen"
d["CreationDate"] = datetime.datetime.today()
def seriel_saving_of_error_rates(cfg, df_plot_sorted, filename, d_fits):
groupby = df_plot_sorted.groupby("tax_id", sort=False, observed=True)
task_id_status_plotting = progress.add_task(
"task_status_plotting",
progress_type="status",
status="Plotting",
name="Plots:",
total=len(groupby),
)
utils.init_parent_folder(filename)
with PdfPages(filename) as pdf, progress:
for tax_id, group in groupby:
fig = plot_single_group(group, cfg, d_fits)
pdf.savefig(fig, bbox_inches="tight", pad_inches=0.1)
progress.advance(task_id_status_plotting)
d = pdf.infodict()
d["Title"] = "Error Rate Distributions"
d["Author"] = "Christian Michelsen"
d["CreationDate"] = datetime.datetime.today()
def save(self, df, metadata=None):
utils.init_parent_folder(self.filename)
table = self._df_to_table_with_metadata(df, metadata)
# pq.write_to_dataset(table, self.filename, partition_cols=partition_cols)
pq.write_table(table, self.filename, version="2.0")