def test_pipeline(self):
df = pd.DataFrame({
'x': np.array(([0.] * 10 + [1] * 10)),
'y': np.array(([0.] * 5 + [1] * 5 + [0] * 5 + [1] * 5)),
'f64': np.arange(20, dtype='f8')
})
df.f64.iloc[2] = np.nan
cvs = ds.Canvas(plot_width=2, plot_height=2, x_range=(0, 1), y_range=(0, 1))
pipeline = ds.Pipeline(df, ds.Point('x', 'y'))
img = pipeline((0, 1), (0, 1), 2, 2)
agg = cvs.points(df, 'x', 'y', ds.count())
self.assertTrue(img.equals(tf.shade(agg)))
color_fn = lambda agg: tf.shade(agg, 'pink', 'red')
pipeline.color_fn = color_fn
img = pipeline((0, 1), (0, 1), 2, 2)
self.assertTrue(img.equals(color_fn(agg)))
transform_fn = lambda agg: agg + 1
pipeline.transform_fn = transform_fn
img = pipeline((0, 1), (0, 1), 2, 2)
self.assertTrue(img.equals(color_fn(transform_fn(agg))))
pipeline = ds.Pipeline(df, ds.Point('x', 'y'), ds.sum('f64'))
img = pipeline((0, 1), (0, 1), 2, 2)
agg = cvs.points(df, 'x', 'y', ds.sum('f64'))
self.assertTrue(img.equals(tf.shade(agg)))
def test_image_initialize():
plt.figure(dpi=100)
ax = plt.subplot(111)
da = dsshow(df, ds.Point("x", "y"), ax=ax)
data = da.get_ds_data()
assert data[0, 0] == 5
assert data[0, -1] == 5
assert data[-1, 0] == 5
assert data[-1, -1] == 5
def test_pipeline():
pipeline = ds.Pipeline(df, ds.Point('x', 'y'))
img = pipeline((0, 1), (0, 1), 2, 2)
agg = cvs.points(df, 'x', 'y', ds.count())
assert img.equals(tf.shade(agg))
color_fn = lambda agg: tf.shade(agg, 'pink', 'red')
pipeline.color_fn = color_fn
img = pipeline((0, 1), (0, 1), 2, 2)
assert img.equals(color_fn(agg))
transform_fn = lambda agg: agg + 1
pipeline.transform_fn = transform_fn
img = pipeline((0, 1), (0, 1), 2, 2)
assert img.equals(color_fn(transform_fn(agg)))
pipeline = ds.Pipeline(df, ds.Point('x', 'y'), ds.sum('f64'))
img = pipeline((0, 1), (0, 1), 2, 2)
agg = cvs.points(df, 'x', 'y', ds.sum('f64'))
assert img.equals(tf.shade(agg))
def correlation_plot(x, y, ax, method):
ax.legend()
ax.grid(True)
ax.set_ylabel(f'Amplitude Aproximada {method}')
ax.set_xlabel('Amplitude de Referencia')
df = pd.DataFrame(dict(x=x, y=y))
da1 = dsshow(df, ds.Point('x', 'y'), norm='log', aspect='auto', ax=ax, x_range=(-100, 400), y_range=(-100, 400), shade_hook=partial(tf.dynspread, threshold=0.8))
fig.colorbar(da1, ax=ax).set_label('Densidade')
return ax
def test_image_update():
fig = plt.figure(dpi=100)
ax = plt.subplot(111)
da = dsshow(df, ds.Point("x", "y"), ax=ax)
ax.set_xlim(0, 0.5)
ax.set_ylim(0, 0.5)
fig.canvas.draw()
data = da.get_ds_data()
assert data[0, 0] == 5
assert data[0, -1] == 0
assert data[-1, 0] == 0
assert data[-1, -1] == 0
def shade_scatter(
dfs,
in_ax=None,
figsize: float = 6,
pixels: int = 1000,
spread_px: int = 1,
spread_threshold: float = 0.2,
min_alpha: int = 10,
color_map=None,
color_key: dict = None,
mask_values: list = None,
mask_name: str = "NA",
mask_color: str = "k",
ax_label_size: float = 12,
frame_offset: float = 0.05,
spine_width: float = 0.5,
spine_color: str = "k",
displayed_sides: tuple = ("bottom", "left"),
legend_ondata: bool = True,
legend_onside: bool = True,
legend_size: float = 12,
legends_per_col: int = 20,
titles: Union[str, List[str]] = None,
title_size: int = 12,
hide_title: bool = False,
cbar_shrink: float = 0.6,
marker_scale: float = 70,
lspacing: float = 0.1,
cspacing: float = 1,
savename: str = None,
dpi: int = 300,
force_ints_as_cats: bool = True,
n_columns: int = 4,
w_pad: float = None,
h_pad: float = None,
show_fig: bool = True,
):
"""
Shows shaded scatter plots. If more then one dataframe is provided it will place the scatterplots in a grid.
"""
import datashader as dsh
from datashader.mpl_ext import dsshow
import datashader.transfer_functions as tf
from functools import partial
titles = _handle_titles_type(titles, len(dfs))
axs = _create_axes(dfs, in_ax, figsize, figsize, w_pad, h_pad, n_columns)
for n, df in _iter_dataframes(dfs, mask_values, mask_name,
force_ints_as_cats):
dim1, dim2, vc = df.columns[:3]
v = df[vc]
col_map, col_key = _scatter_make_colors(v, color_map, color_key,
mask_color, mask_name)
if v.dtype.name == "category":
agg = dsh.count_cat(vc)
else:
if v.nunique() == 1:
agg = dsh.count(vc)
else:
agg = dsh.mean(vc)
ax = axs[int(n / n_columns), n % n_columns]
artist = dsshow(
df,
dsh.Point(dim1, dim2),
aggregator=agg,
norm="eq_hist",
color_key=col_key,
cmap=col_map,
alpha_range=(min_alpha, 255),
shade_hook=partial(tf.dynspread,
threshold=spread_threshold,
max_px=spread_px),
plot_height=pixels,
plot_width=pixels,
aspect="equal",
width_scale=1,
height_scale=1,
ax=ax,
)
_scatter_label_axis(df, ax, ax_label_size, frame_offset)
_scatter_cleanup(ax, spine_width, spine_color, displayed_sides)
if titles is not None:
title = titles[n]
else:
title = None
_scatter_legends(
df,
ax,
col_map,
col_key,
legend_ondata,
legend_onside,
legend_size,
title,
title_size,
hide_title,
legends_per_col,
marker_scale,
lspacing,
cspacing,
cbar_shrink,
)
if savename:
plt.savefig(savename, dpi=dpi, bbox_inches="tight")
if show_fig:
plt.show()
else:
return axs
def shade_scatter(dfs,
in_ax=None,
figsize: float = 6,
pixels: int = 1000,
sampling: float = 0.1,
spread_px: int = 1,
spread_threshold: float = 0.2,
min_alpha: int = 10,
color_map=None,
color_key: dict = None,
mask_values: list = None,
mask_name: str = 'NA',
mask_color: str = 'k',
ax_label_size: float = 12,
frame_offset: float = 0.05,
spine_width: float = 0.5,
spine_color: str = 'k',
displayed_sides: tuple = ('bottom', 'left'),
legend_ondata: bool = True,
legend_onside: bool = True,
legend_size: float = 12,
legends_per_col: int = 20,
cbar_shrink: float = 0.6,
marker_scale: float = 70,
lspacing: float = 0.1,
cspacing: float = 1,
savename: str = None,
dpi: int = 300,
force_ints_as_cats: bool = True,
n_columns: int = 4,
w_pad: float = None,
h_pad: float = None,
show_fig: bool = True):
"""
Shows shaded scatter plots. If more then one dataframe is provided it will place the scatterplots in a grid.
"""
import datashader as dsh
from IPython.display import display
from datashader.mpl_ext import dsshow
import datashader.transfer_functions as tf
from functools import partial
if len(dfs) > 1:
if in_ax is not None:
logger.warning(
f"'in_ax' will not be used as multiple attributes will be plotted. Using internal grid"
f"layout")
fig, axs = _make_grid(figsize, figsize, w_pad, h_pad, len(dfs),
n_columns)
else:
if in_ax is None:
_, axs = plt.subplots(1,
1,
figsize=(figsize, figsize),
squeeze=False)
else:
axs = in_ax
for i, df in enumerate(dfs):
dim1, dim2, vc = df.columns[:3]
v = _scatter_fix_mask(df[vc].copy(), mask_values, mask_name)
v = _scatter_fix_type(v, force_ints_as_cats)
df[vc] = v
col_map, col_key = _scatter_make_colors(v, color_map, color_key,
mask_color, mask_name)
if v.dtype.name == 'category':
agg = dsh.count_cat(vc)
else:
if v.nunique() == 1:
agg = dsh.count(vc)
else:
agg = dsh.mean(vc)
ax = axs[int(i / n_columns), i % n_columns]
artist = dsshow(df,
dsh.Point(dim1, dim2),
aggregator=agg,
norm='eq_hist',
color_key=col_key,
cmap=col_map,
alpha_range=(min_alpha, 255),
shade_hook=partial(tf.dynspread,
threshold=spread_threshold,
max_px=spread_px),
plot_height=pixels,
plot_width=pixels,
aspect='equal',
width_scale=1,
height_scale=1,
ax=ax)
_scatter_label_axis(df, ax, ax_label_size, frame_offset)
_scatter_cleanup(ax, spine_width, spine_color, displayed_sides)
_scatter_legends(df, ax, col_map, col_key, legend_ondata,
legend_onside, legend_size, legends_per_col,
marker_scale, lspacing, cspacing, cbar_shrink)
if savename:
plt.savefig(savename, dpi=dpi, bbox_inches='tight')
if show_fig:
plt.show()
else:
return axs