transforms Module

Various transforms for FCS data

flowutils.transforms.logicle(data, channel_indices, t=262144, m=4.5, w=0.5, a=0)

Logicle transformation, implemented as defined in the GatingML 2.0 specification:

logicle(x, T, W, M, A) = root(B(y, T, W, M, A) − x)

where B is a modified bi-exponential function defined as:

B(y, T, W, M, A) = ae^(by) − ce^(−dy) − f

The Logicle transformation was originally defined in the publication:

Moore WA and Parks DR. Update for the logicle data scale including operational code implementations. Cytometry A., 2012:81A(4):273–277.

Parameters:

data – NumPy array of FCS event data. If a 1-D array, channel_indices option is ignored
channel_indices – channel indices to transform (other channels returned in place, untransformed). If None, then all events will be transformed.
t – parameter for the top of the linear scale (e.g. 262144)
m – parameter for the number of decades the true logarithmic scale approaches at the high end of the scale
w – parameter for the approximate number of decades in the linear region
a – parameter for the additional number of negative decades

Returns:

NumPy array of transformed events

flowutils.transforms.logicle_inverse(data, channel_indices, t=262144, m=4.5, w=0.5, a=0)

Inverse of the Logicle transformation (see logicle() documentation for more details)

Parameters:

data – NumPy array of FCS event data. If a 1-D array, channel_indices option is ignored
channel_indices – channel indices to transform (other channels returned in place, untransformed). If None, then all events will be transformed.
t – parameter for the top of the linear scale (e.g. 262144)
m – parameter for the number of decades the true logarithmic scale approaches at the high end of the scale
w – parameter for the approximate number of decades in the linear region
a – parameter for the additional number of negative decades

Returns:

NumPy array of transformed events

flowutils.transforms.hyperlog(data, channel_indices, t=262144, m=4.5, w=0.5, a=0)

Hyperlog transformation, implemented as defined in the GatingML 2.0 specification:

hyperlog(x, T, W, M, A) = root(EH(y, T, W, M, A) − x)

where EH is defined as:

EH(y, T, W, M, A) = ae^(by) + cy − f

The Hyperlog transformation was originally defined in the publication:

Bagwell CB. Hyperlog-a flexible log-like transform for negative, zero, and positive valued data. Cytometry A., 2005:64(1):34–42.

Parameters:

data – NumPy array of FCS event data. If a 1-D array, channel_indices option is ignored
channel_indices – channel indices to transform (other channels returned in place, untransformed). If None, then all events will be transformed.
t – parameter for the top of the linear scale (e.g. 262144)
m – parameter for desired number of decades
w – parameter for the approximate number of decades in the linear region
a – parameter for the additional number of negative decades

Returns:

NumPy array of transformed events

flowutils.transforms.hyperlog_inverse(data, channel_indices, t=262144, m=4.5, w=0.5, a=0)

Inverse of the Hyperlog transformation, implemented as defined in the GatingML 2.0 specification (see hyperlog() documentation for more details).

Parameters:

data – NumPy array of FCS event data. If a 1-D array, channel_indices option is ignored
channel_indices – channel indices to transform (other channels returned in place, untransformed). If None, then all events will be transformed.
t – parameter for the top of the linear scale (e.g. 262144)
m – parameter for desired number of decades
w – parameter for the approximate number of decades in the linear region
a – parameter for the additional number of negative decades

Returns:

NumPy array of transformed events

flowutils.transforms.asinh(data, channel_indices, t, m, a)

An implementation of the parametrized inverse hyperbolic sine function as defined in the GatingML 2.0 specification.

Parameters:

data – NumPy array of FCS event data. If a 1-D array, channel_indices option is ignored
channel_indices – channel indices to transform (other channels returned in place, untransformed). If None, then all events will be transformed.
t – parameter specifying the top of the scale, (e.g. 262144)
m – parameter for the number of decades
a – parameter for the number of additional negative decades

Returns:

NumPy array of transformed events

flowutils.transforms.asinh_inverse(data, channel_indices, t, m, a)

Inverse of the parametrized inverse hyperbolic sine function as defined in the GatingML 2.0 specification.

Parameters:

data – NumPy array of FCS event data. If a 1-D array, channel_indices option is ignored
channel_indices – channel indices to transform (other channels returned in place, untransformed). If None, then all events will be transformed.
t – parameter specifying the top of the scale, (e.g. 262144)
m – parameter for the number of decades
a – parameter for the number of additional negative decades

Returns:

NumPy array of transformed events

flowutils.transforms.log(data, channel_indices, t, m)

Parametrized logarithmic transformation, implemented as defined in the GatingML 2.0 specification:

flog(x, T, M) = (1 / M) * log_10(x / T) + 1

Parameters:

data – NumPy array of FCS event data. If a 1-D array, channel_indices option is ignored
channel_indices – channel indices to transform (other channels returned in place, untransformed). If None, then all events will be transformed.
t – parameter for the top of the linear scale (e.g. 262144)
m – parameter for desired number of decades

Returns:

NumPy array of transformed events

flowutils.transforms.log_inverse(data, channel_indices, t, m)

Inverse of logarithmic transformation

Parameters:

data – NumPy array of FCS event data. If a 1-D array, channel_indices option is ignored
channel_indices – channel indices to transform (other channels returned in place, untransformed). If None, then all events will be transformed.
t – parameter for the top of the linear scale (e.g. 262144)
m – parameter for desired number of decades

Returns:

NumPy array of transformed events