reda.plotters package

Various tools for data visualization

Submodules

reda.plotters.histograms module

Histogram functions for raw data.

reda.plotters.histograms.plot_histograms(ertobj, keys, **kwargs)

Generate histograms for one or more keys in the given container.

Parameters:

ertobjcontainer instance or pandas.DataFrame

data object which contains the data.

keysstr or list of strings

which keys (column names) to plot

mergebool, optional

if True, then generate only one figure with all key-plots as columns (default True)

log10plotbool, optional

default: True

extra_dimslist, optional

?

nr_binsNone|int

if an int is given, use this as the number of bins, otherwise use a heuristic.

Returns:

figuresdict

dictionary with the generated histogram figures

Examples

>>> from reda.plotters import plot_histograms
>>> from reda.testing.containers import get_simple_ert_container_nor
>>> ERTContainer = get_simple_ert_container_nor()
>>> figs_dict = plot_histograms(ERTContainer, "r", merge=False)

reda.plotters.histograms.plot_histograms_extra_dims(dataobj, keys, primary_dim=None, **kwargs)

Produce histograms grouped by one extra dimensions. Produce additional figures for all extra dimensions.

The dimension to spread out along subplots is called the primary extra dimension.

Extra dimensions are:

• timesteps

• frequency

If only “timesteps” are present, timesteps will be plotted as subplots.

If only “frequencies” are present, frequencies will be plotted as subplots.

If more than one extra dimensions is present, multiple figures will be generated.

Test cases:

• not extra dims present (primary_dim=None, timestep, frequency)

• timestep (primary_dim=None, timestep, frequency)

• frequency (primary_dim=None, timestep, frequency)

• timestep + frequency (primary_dim=None, timestep, frequency)

check nr of returned figures.

Parameters:

dataobjpandas.DataFrame or reda container

The data container/data frame which holds the data

keysstring|list|tuple|iterable

The keys (columns) of the dataobj to plot

primary_dimstring

primary extra dimension to plot along subplots. If None, the first extra dimension found in the data set is used, in the following order: timestep, frequency.

subquerystr, optional

if provided, apply this query statement to the data before plotting

log10bool

Plot only log10 transformation of data (default: False)

lin_and_log10bool

Plot both linear and log10 representation of data (default: False)

Nxint, optional

Number of subplots in x direction

Returns:

dim_namestr

name of secondary dimensions, i.e. the dimensions for which separate figures were created

figuresdict

dict containing the generated figures. The keys of the dict correspond to the secondary dimension grouping keys

Examples

>>> from reda.testing.containers import get_simple_ert_container_norrec
>>> ert = get_simple_ert_container_norrec()
>>> import reda.plotters.histograms as RH
>>> dim_name, fig = RH.plot_histograms_extra_dims(ert, ['r', ])

>>> from reda.testing.containers import get_simple_ert_container_norrec
>>> ert = get_simple_ert_container_norrec()
>>> import reda.plotters.histograms as RH
>>> dim_name, fig = RH.plot_histograms_extra_dims(ert, ['r', 'a'])

reda.plotters.plots2d module

2D plots of raw data. This includes pseudosections.

reda.plotters.plots2d.matplot(x, y, z, ax=None, colorbar=True, **kwargs)

Plot x, y, z as expected with correct axis labels.

Examples

>>> import matplotlib.pyplot as plt
>>> import numpy as np
>>> from reda.plotters import matplot
>>> a = np.arange(4)
>>> b = np.arange(3) + 3
>>> def sum(a, b):
...    return a + b
>>> x, y = np.meshgrid(a, b)
>>> c = sum(x, y)
>>> fig, (ax1, ax2) = plt.subplots(1, 2)
>>> im = ax1.pcolormesh(x, y, c, shading='auto')
>>> _ = plt.colorbar(im, ax=ax1)
>>> _ = ax1.set_title("plt.pcolormesh")
>>> _, _ = matplot(x, y, c, ax=ax2)
>>> _ = ax2.set_title("reda.plotters.matplot")
>>> fig.show()

(Source code, png, pdf)

reda.plotters.plots2d.plot_pseudodepths(configs, nr_electrodes, spacing=1, grid=None, ctypes=None, dd_merge=False, **kwargs)

Plot pseudodepths for the measurements. If grid is given, then the actual electrode positions are used, and the parameter ‘spacing’ is ignored’

Parameters:

configs: :class:`numpy.ndarray`

Nx4 array containing the quadrupoles for different measurements

nr_electrodes: int

The overall number of electrodes of the dataset. This is used to plot the surface electrodes

spacing: float, optional

assumed distance between electrodes. Default=1

grid: crtomo.grid.crt_grid instance, optional

grid instance. Used to infer real electrode positions

ctypes: list of strings, optional

a list of configuration types that will be plotted. All configurations that can not be sorted into these types will not be plotted! Possible types:

• dd

• schlumberger

dd_merge: bool, optional

if True, merge all skips. Otherwise, generate individual plots for each skip

Returns:

figs: matplotlib.figure.Figure instance or list of Figure instances

if only one type was plotted, then the figure instance is returned. Otherwise, return a list of figure instances.

axes: axes object or list of axes ojects

plot axes

Examples

from reda.plotters.plots2d import plot_pseudodepths
# define a few measurements
import numpy as np
configs = np.array((
    (1, 2, 4, 3),
    (1, 2, 5, 4),
    (1, 2, 6, 5),
    (2, 3, 5, 4),
    (2, 3, 6, 5),
    (3, 4, 6, 5),
))
# plot
fig, axes = plot_pseudodepths(configs, nr_electrodes=6, spacing=1,
                              ctypes=['dd', ])

(Source code, png, pdf)

from reda.plotters.plots2d import plot_pseudodepths
# define a few measurements
import numpy as np
configs = np.array((
    (4, 7, 5, 6),
    (3, 8, 5, 6),
    (2, 9, 5, 6),
    (1, 10, 5, 6),
))
# plot
fig, axes = plot_pseudodepths(configs, nr_electrodes=10, spacing=1,
                              ctypes=['schlumberger', ])

(Source code, png, pdf)

reda.plotters.plots2d.plot_pseudosection(df, plot_key, spacing=1, ctypes=None, dd_merge=False, cb=False, **kwargs)

Create a pseudosection plot for a given measurement

Parameters:

df: dataframe

measurement dataframe, one measurement frame (i.e., only one frequency etc)

key:

which key to colorcode

spacing: float, optional

assumed electrode spacing

ctypes: list of strings

which configurations to plot, default: dd

dd_merge: bool, optional

?

cb: bool, optional

?

reda.plotters.plots2d.plot_rawdataplot()

reda.plotters.pseudoplots module

reda.plotters.pseudoplots.plot_ps_extra(dataobj, key, **kwargs)

Create grouped pseudoplots for one or more time steps

Parameters:

dataobj: :class:`reda.containers.ERT`

An ERT container with loaded data

key: string

The column name to plot

subquery: string, optional

cbmin: float, optional

cbmax: float, optional

Examples

>>> from reda.testing.containers import get_simple_ert_container_norrec
>>> ert = get_simple_ert_container_norrec()
>>> import reda.plotters.pseudoplots as PS
>>> fig = PS.plot_ps_extra(ert, key='r')

reda.plotters.pseudoplots.plot_pseudosection_type1(dataobj, column, **kwargs)

Create a pseudosection plot of type 1.

For a given measurement data set, create plots that graphically show the data in a 2D color plot. Hereby, x and y coordinates in the plot are determined by the current dipole (x-axis) and voltage dipole (y-axis) of the corresponding measurement configurations.

This type of rawdata plot can plot any type of measurement configurations, i.e., it is not restricted to certain types of configurations such as Dipole-dipole or Wenner configurations. However, spatial inferences cannot be made from the plots for all configuration types.

Coordinates are generated by separately sorting the dipoles (current/voltage) along the first electrode, and then subsequently sorting by the difference (skip) between both electrodes.

Note that this type of raw data plot does not take into account the real electrode spacing of the measurement setup.

Type 1 plots can show normal and reciprocal data at the same time. Hereby the upper left triangle of the plot area usually contains normal data, and the lower right triangle contains the corresponding reciprocal data. Therefore a quick assessment of normal-reciprocal differences can be made by visually comparing the symmetry on the 1:1 line going from the lower left corner to the upper right corner.

Note that this interpretation usually only holds for Dipole-Dipole data (and the related Wenner configurations).

Parameters:

dataobj: ERT container|pandas.DataFrame

Container or DataFrame with data to plot

column: string

Column key to plot

ax: matplotlib.Axes object, optional

axes object to plot to. If not provided, a new figure and axes object will be created and returned

nocb: bool, optional

if set to False, don’t plot the colorbar

cblabel: string, optional

label for the colorbar

cbmin: float, optional

colorbar minimum

cbmax: float, optional

colorbar maximum

xlabel: string, optional

xlabel for the plot

ylabel: string, optional

ylabel for the plot

do_not_saturate: bool, optional

if set to True, then values outside the colorbar range will not saturate with the respective limit colors. Instead, values lower than the CB are colored “cyan” and vaues above the CB limit are colored “red”

log10: bool, optional

if set to True, plot the log10 values of the provided data

Returns:

fig:

figure object

ax:

axes object

cb:

colorbar object

Examples

You can just supply a pandas.DataFrame to the plot function:

import numpy as np
configs = np.array((
    (1, 2, 4, 3),
    (1, 2, 5, 4),
    (1, 2, 6, 5),
    (2, 3, 5, 4),
    (2, 3, 6, 5),
    (3, 4, 6, 5),
))
measurements = np.random.random(configs.shape[0])
import pandas as pd
df = pd.DataFrame(configs, columns=['a', 'b', 'm', 'n'])
df['measurements'] = measurements

from reda.plotters.pseudoplots import plot_pseudosection_type2
fig, ax, cb = plot_pseudosection_type2(
   dataobj=df,
   column='measurements',
)

(Source code, png, pdf)

You can also supply axes to plot to:

import numpy as np
configs = np.array((
    (1, 2, 4, 3),
    (1, 2, 5, 4),
    (1, 2, 6, 5),
    (2, 3, 5, 4),
    (2, 3, 6, 5),
    (3, 4, 6, 5),
))
measurements = np.random.random(configs.shape[0])
measurements2 = np.random.random(configs.shape[0])

import pandas as pd
df = pd.DataFrame(configs, columns=['a', 'b', 'm', 'n'])
df['measurements'] = measurements
df['measurements2'] = measurements2

from reda.plotters.pseudoplots import plot_pseudosection_type2

fig, axes = plt.subplots(1, 2)

plot_pseudosection_type2(
    df,
    column='measurements',
    ax=axes[0],
    cblabel='this label',
    xlabel='xlabel',
    ylabel='ylabel',
)
plot_pseudosection_type2(
    df,
    column='measurements2',
    ax=axes[1],
    cblabel='measurement 2',
    xlabel='xlabel',
    ylabel='ylabel',
)
fig.tight_layout()

(Source code, png, pdf)

>>> from reda.testing.containers import get_simple_ert_container_norrec
>>> ERTContainer_nr = get_simple_ert_container_norrec()
>>> import reda.plotters.pseudoplots as ps
>>> fig, axes, cb = ps.plot_pseudosection_type2(ERTContainer_nr, 'r')

reda.plotters.pseudoplots.plot_pseudosection_type2(dataobj, column, **kwargs)

Create a pseudosection plot of type 2.

For a given measurement data set, create plots that graphically show the data in a 2D pseudoplot. Hereby, x and y coordinates (pseudodistance and pseudodepth) in the plot are determined by the corresponding measurement configuration (after Roy and Apparao (1971) and Dahlin and Zou (2005)).

This type of rawdata plot can plot any type of measurement configurations, i.e., it is not restricted to certain types of configurations such as Dipole-dipole or Wenner configurations.

Parameters:

dataobj: ERT container|pandas.DataFrame

Container or DataFrame with data to plot

column: string

Column key to plot

ax: matplotlib.Axes object, optional

axes object to plot to. If not provided, a new figure and axes object will be created and returned

nocb: bool, optional

if set to False, don’t plot the colorbar

cblabel: string, optional

label for the colorbar

cbmin: float, optional

colorbar minimum

cbmax: float, optional

colorbar maximum

xlabel: string, optional

xlabel for the plot

ylabel: string, optional

ylabel for the plot

do_not_saturate: bool, optional

if set to True, then values outside the colorbar range will not saturate with the respective limit colors. Instead, values lower than the CB are colored “cyan” and vaues above the CB limit are colored “red”

markersize: float, optional

size of plotted data points

spacing: float, optional

if set to True, the actual electrode spacing is used for the computation of the pseudodepth and -distance; default value is 1 m

log10: bool, optional

if set to True, plot the log10 values of the provided data

force_to_normalbool, optional (False)

If True, force all data points to be plotted below the y=0 axis. This is useful to visualize nor-rec-merged data sets.

Returns:

fig:

figure object

ax:

axes object

cb:

colorbar object

Examples

You can just supply a pandas.DataFrame to the plot function:

import numpy as np
configs = np.array((
    (1, 2, 4, 3),
    (1, 2, 5, 4),
    (1, 2, 6, 5),
    (2, 3, 5, 4),
    (2, 3, 6, 5),
    (3, 4, 6, 5),
))
measurements = np.random.random(configs.shape[0])
import pandas as pd
df = pd.DataFrame(configs, columns=['a', 'b', 'm', 'n'])
df['measurements'] = measurements

from reda.plotters.pseudoplots import plot_pseudosection_type2
fig, ax, cb = plot_pseudosection_type2(
   dataobj=df,
   column='measurements',
)

(Source code, png, pdf)

You can also supply axes to plot to:

import numpy as np
configs = np.array((
    (1, 2, 4, 3),
    (1, 2, 5, 4),
    (1, 2, 6, 5),
    (2, 3, 5, 4),
    (2, 3, 6, 5),
    (3, 4, 6, 5),
))
measurements = np.random.random(configs.shape[0])
measurements2 = np.random.random(configs.shape[0])

import pandas as pd
df = pd.DataFrame(configs, columns=['a', 'b', 'm', 'n'])
df['measurements'] = measurements
df['measurements2'] = measurements2

from reda.plotters.pseudoplots import plot_pseudosection_type2

fig, axes = plt.subplots(1, 2)

plot_pseudosection_type2(
    df,
    column='measurements',
    ax=axes[0],
    cblabel='this label',
    xlabel='xlabel',
    ylabel='ylabel',
)
plot_pseudosection_type2(
    df,
    column='measurements2',
    ax=axes[1],
    cblabel='measurement 2',
    xlabel='xlabel',
    ylabel='ylabel',
)
fig.tight_layout()

(Source code, png, pdf)

reda.plotters.pseudoplots.test_get_unique_identifiers(c)

reda.plotters.pseudoplots_type_3_crtomo module

Sensitivity-based pseudoplots

reda.plotters.pseudoplots_type_3_crtomo.plot_pseudosection_type3(dataobj, column, log10, crmod_settings, **kwargs)

settings = {

‘rho’: 100, # resistivity to use for homogeneous model, [Ohm m] ‘elem’ ‘elec’ ‘2D’ : True|False ‘sink_node’: ‘100’,

}

reda.plotters.time_series module

reda.plotters.time_series.plot_quadpole_evolution(dataobj, quadpole, cols, threshold=5, rolling=False, ax=None)

Visualize time-lapse evolution of a single quadropole.

Parameters:

dataobjpandas.DataFrame

DataFrame containing the data. Please refer to the documentation for required columns.

quadpolelist of integers

Electrode numbers of the the quadropole.

colsstr

The column/parameter to plot over time.

thresholdfloat

Allowed percentage deviation from the rolling standard deviation.

rollingbool

Calculate rolling median values (the default is False).

axmpl.axes

Optional axes object to plot to.