Skip to content

Commit

Permalink
Readers now return dictionaries with instance names
Browse files Browse the repository at this point in the history
  • Loading branch information
@hrayrhar
hrayrhar committed Jan 18, 2018
1 parent a4e331e commit b37f83d
Showing 1 changed file with 63 additions and 26 deletions.
89 changes: 63 additions & 26 deletions mimic3benchmark/readers.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -74,28 +74,34 @@ def read_example(self, index):
index : int
Index of the line of the listfile to read (counting starts from 0).
Returns (X, t, y, header)
Returns dictionary
-------
X : np.array
2D array containing all events. Each row corresponds to a moment.
First coloumn is the time and other columns correspond to different
First column is the time and other columns correspond to different
variables.
t : float
Lenght of the data in hours. Note, in general, it is not eqaul to the
Length of the data in hours. Note, in general, it is not equal to the
timestamp of last event.
y : int (0 or 1)
Mortality within next 24 hours.
header : array of strings
Names of the columns. The ordering of the columns is always the same.
name: Name of the sample.
"""
if (index < 0 or index >= len(self._data)):
raise ValueError("Index must be from 0 (inclusive) to number of examples (exclusive).")

name = self._data[index][0]
t = self._data[index][1]
(X, header) = self._read_timeseries(self._data[index][0], t)
y = self._data[index][2]
(X, header) = self._read_timeseries(name, t)

return (X, t, y, header)
return {"X": X,
"t": t,
"y": y,
"header": header,
"name": name}


class InHospitalMortalityReader(Reader):
Expand Down Expand Up @@ -138,27 +144,34 @@ def read_example(self, index):
index : int
Index of the line of the listfile to read (counting starts from 0).
Returns (X, t, y, header)
Returns dictionary
-------
X : np.array
2D array containing all events. Each row corresponds to a moment.
First coloumn is the time and other columns correspond to different
First column is the time and other columns correspond to different
variables.
t : float
Lenght of the data in hours. Note, in general, it is not eqaul to the
Length of the data in hours. Note, in general, it is not equal to the
timestamp of last event.
y : int (0 or 1)
In-hospital mortality.
header : array of strings
Names of the columns. The ordering of the columns is always the same.
name: Name of the sample.
"""
if (index < 0 or index >= len(self._data)):
raise ValueError("Index must be from 0 (inclusive) to number of lines (exclusive).")

(X, header) = self._read_timeseries(self._data[index][0])
name = self._data[index][0]
t = self._period_length
y = self._data[index][1]
(X, header) = self._read_timeseries(name)

return (X, self._period_length, y, header)
return {"X": X,
"t": t,
"y": y,
"header": header,
"name": name}


class LengthOfStayReader(Reader):
Expand Down Expand Up @@ -201,28 +214,34 @@ def read_example(self, index):
index : int
Index of the line of the listfile to read (counting starts from 0).
Returns (X, t, y, header)
Returns dictionary
-------
X : np.array
2D array containing all events. Each row corresponds to a moment.
First coloumn is the time and other columns correspond to different
First column is the time and other columns correspond to different
variables.
t : float
Lenght of the data in hours. Note, in general, it is not eqaul to the
Length of the data in hours. Note, in general, it is not equal to the
timestamp of last event.
y : float
Remaining time in ICU.
header : array of strings
Names of the columns. The ordering of the columns is always the same.
name: Name of the sample.
"""
if (index < 0 or index >= len(self._data)):
raise ValueError("Index must be from 0 (inclusive) to number of lines (exclusive).")


name = self._data[index][0]
t = self._data[index][1]
(X, header) = self._read_timeseries(self._data[index][0], t)
y = self._data[index][2]
(X, header) = self._read_timeseries(name, t)

return (X, t, y, header)
return {"X": X,
"t": t,
"y": y,
"header": header,
"name": name}


class PhenotypingReader(Reader):
Expand Down Expand Up @@ -264,27 +283,34 @@ def read_example(self, index):
index : int
Index of the line of the listfile to read (counting starts from 0).
Returns (X, t, y, header)
Returns dictionary
-------
X : np.array
2D array containing all events. Each row corresponds to a moment.
First coloumn is the time and other columns correspond to different
First column is the time and other columns correspond to different
variables.
t : float
Lenght of the data in hours. Note, in general, it is not eqaul to the
Length of the data in hours. Note, in general, it is not equal to the
timestamp of last event.
y : array of ints
Phenotype labels.
header : array of strings
Names of the columns. The ordering of the columns is always the same.
name: Name of the sample.
"""
if (index < 0 or index >= len(self._data)):
raise ValueError("Index must be from 0 (inclusive) to number of lines (exclusive).")

(X, header) = self._read_timeseries(self._data[index][0])
name = self._data[index][0]
t = self._data[index][1]
y = self._data[index][2]
(X, header) = self._read_timeseries(name)

return (X, self._data[index][1], y, header)
return {"X": X,
"t": t,
"y": y,
"header": header,
"name": name}


class MultitaskReader(Reader):
Expand Down Expand Up @@ -343,14 +369,14 @@ def read_example(self, index):
index : int
Index of the line of the listfile to read (counting starts from 0).
Returns (X, t, ihm, los, ph, decomp, header)
Returns dictionary
-------
X : np.array
2D array containing all events. Each row corresponds to a moment.
First coloumn is the time and other columns correspond to different
First column is the time and other columns correspond to different
variables.
t : float
Lenght of the data in hours. Note, in general, it is not eqaul to the
Length of the data in hours. Note, in general, it is not equal to the
timestamp of last event.
ihm : array
Array of 3 integers: [pos, mask, label].
Expand All @@ -362,8 +388,19 @@ def read_example(self, index):
Array of 2 arrays: [masks, labels].
header : array of strings
Names of the columns. The ordering of the columns is always the same.
name: Name of the sample.
"""
if (index < 0 or index >= len(self._data)):
raise ValueError("Index must be from 0 (inclusive) to number of lines (exclusive).")
(X, header) = self._read_timeseries(self._data[index][0])
return [X] + list(self._data[index][1:]) + [header]

name = self._data[index][0]
(X, header) = self._read_timeseries(name)

return {"X": X,
"t": self.data[index][1],
"ihm": self.data[index][2],
"los": self.data[index][3],
"ph": self.data[index][4],
"decomp": self.data[index][5],
"header": header,
"name": name}

0 comments on commit b37f83d

Please sign in to comment.