diff --git a/src/agora/io/signal.py b/src/agora/io/signal.py
index 3a5261f7bc7f02374f2f10980d271afbfaa2f104..4193178abb71a2ca81d03c7435725ef28ef2fc57 100644
--- a/src/agora/io/signal.py
+++ b/src/agora/io/signal.py
@@ -59,7 +59,7 @@ class Signal(BridgeH5):
):
"""Get Signal after merging and picking."""
if isinstance(dset_name, str):
- dsets = self.get_raw(dset_name, tmax_in_mins)
+ dsets = self.get_raw(dset_name, tmax_in_mins=tmax_in_mins)
if dsets is not None:
picked_merged = self.apply_merging_picking(dsets)
return self.add_name(picked_merged, dset_name)
@@ -76,10 +76,7 @@ class Signal(BridgeH5):
def cols_in_mins(self, df: pd.DataFrame):
"""Convert numerical columns in a data frame to minutes."""
- try:
- df.columns = (df.columns * self.tinterval // 60).astype(int)
- except Exception as e:
- self._log(f"Unable to convert columns to minutes: {e}", "debug")
+ df.columns = (df.columns * self.tinterval // 60).astype(int)
return df
@cached_property
@@ -105,7 +102,7 @@ class Signal(BridgeH5):
)
return 300
- def retained(self, signal, cutoff=0, tmax_in_mins: int = None):
+ def retained(self, signal, cutoff: float = 0, tmax_in_mins: int = None):
"""Get retained cells for a Signal or list of Signals."""
if isinstance(signal, str):
# get data frame
@@ -284,8 +281,12 @@ class Signal(BridgeH5):
if in_minutes:
df = self.cols_in_mins(df)
# limit data by time and discard NaNs
- if tmax_in_mins and type(tmax_in_mins) is int:
- df = df[df.columns[df.columns < tmax_in_mins]]
+ if (
+ in_minutes
+ and tmax_in_mins
+ and type(tmax_in_mins) is int
+ ):
+ df = df[df.columns[df.columns <= tmax_in_mins]]
df = df.dropna(how="all")
# add mother label to data frame
if lineage:
@@ -303,7 +304,7 @@ class Signal(BridgeH5):
df = add_index_levels(
df, {"mother_label": mother_label}
)
- return df
+ return df
elif isinstance(dataset, list):
return [
self.get_raw(
@@ -315,7 +316,7 @@ class Signal(BridgeH5):
for dset in dataset
]
except Exception as e:
- message = f"Signal could not obtain data {dataset}: {e}."
+ message = f"Signal could not obtain data {dataset}: {e}"
self._log(message)
def load_merges(self):
diff --git a/src/postprocessor/grouper.py b/src/postprocessor/grouper.py
index f964af7f58797e20d0529726d89e0e03be7c2a2c..7cf22be19fd7d3767223d88beb82204feaf39807 100644
--- a/src/postprocessor/grouper.py
+++ b/src/postprocessor/grouper.py
@@ -77,6 +77,7 @@ class Grouper(ABC):
pool: t.Optional[int] = None,
mode: str = "retained",
selected_positions: t.List[str] = None,
+ tmax_in_mins_dict: dict = None,
**kwargs,
):
"""
@@ -88,7 +89,7 @@ class Grouper(ABC):
----------
path : str
Signal location within h5 file.
- pool : int
+ pool : int (optional)
Number of threads used; if 0 or None only one core is used.
mode: str
If "retained" (default), return Signal with merging, picking, and lineage
@@ -102,8 +103,13 @@ class Grouper(ABC):
identified mother.
If "families", get Signal with merging, picking, and lineage
information applied.
- selected_positions: list[str]
+ selected_positions: list[str] (optional)
If defined, get signals for only these positions.
+ tmax_in_mins_dict: dict (optional)
+ A dictionary with positions as keys and maximum times in minutes as
+ values. For example: { "PDR5_GFP_001": 6 * 60}.
+ Data will only be include up to this time point, which is a way to
+ avoid errors in assigning lineages because of clogging.
**kwargs : key, value pairings
Named arguments for concat_ind_function
@@ -127,6 +133,7 @@ class Grouper(ABC):
f=concat_one_signal,
pool=pool,
positions=good_positions,
+ tmax_in_mins_dict=tmax_in_mins_dict,
**kwargs,
)
# check for errors
@@ -171,6 +178,7 @@ class Grouper(ABC):
f: t.Callable,
pool: t.Optional[int] = None,
positions: t.Dict[str, Signal] = None,
+ tmax_in_mins_dict: dict = None,
**kwargs,
):
"""
@@ -187,6 +195,7 @@ class Grouper(ABC):
position=x[1],
group=self.positions_groups[x[0]],
position_name=x[0],
+ tmax_in_mins_dict=tmax_in_mins_dict,
**kwargs,
),
positions.items(),
@@ -198,6 +207,7 @@ class Grouper(ABC):
position=position,
group=self.positions_groups[name],
position_name=name,
+ tmax_in_mins_dict=tmax_in_mins_dict,
**kwargs,
)
for name, position in positions.items()
@@ -280,7 +290,7 @@ def concat_one_signal(
mode: str = "retained",
position_name=None,
tmax_in_mins_dict=None,
- **kwargs,
+ cutoff: float = 0,
) -> pd.DataFrame:
"""Retrieve a signal for one position."""
if tmax_in_mins_dict and position_name in tmax_in_mins_dict:
@@ -297,20 +307,24 @@ def concat_one_signal(
else:
print(f" Loading {path} for {position_name}.")
if mode == "retained":
- combined = position.retained(path, tmax_in_mins=tmax_in_mins, **kwargs)
+ # applies picking and merging via Signal.get
+ combined = position.retained(
+ path, tmax_in_mins=tmax_in_mins, cutoff=cutoff
+ )
elif mode == "raw":
- combined = position.get_raw(path, tmax_in_mins=tmax_in_mins, **kwargs)
- elif mode == "daughters":
+ # no picking and merging
+ combined = position.get_raw(path, tmax_in_mins=tmax_in_mins)
+ elif mode == "raw_daughters":
combined = position.get_raw(
- path, lineage=True, tmax_in_mins=tmax_in_mins, **kwargs
+ path, lineage=True, tmax_in_mins=tmax_in_mins
)
if combined is not None:
combined = combined.loc[
combined.index.get_level_values("mother_label") > 0
]
- elif mode == "mothers":
+ elif mode == "raw_mothers":
combined = position.get_raw(
- path, lineage=True, tmax_in_mins=tmax_in_mins, **kwargs
+ path, lineage=True, tmax_in_mins=tmax_in_mins
)
if combined is not None:
combined = combined.loc[
@@ -318,6 +332,7 @@ def concat_one_signal(
]
combined = combined.droplevel("mother_label")
elif mode == "families":
+ # applies picking and merging via Signal.__getitem__
combined = position[path]
else:
raise Exception(f"concat_one_signal: {mode} not recognised.")