def pairplot(self, argin):
"""plot array of plots for all pairs of columns
<query> [<options>]
- continuous-continuous as scatter (optional KDE contour)
- continuous-categorical as violinplot
- categorical-categorical as heatmap
Options:
-f, --filename: the output filename. If not specified, tries
to draw.
-g, --generator: the generator name. Providing the generator
name make help to plot more intelligently.
-s, --shortnames: Use column short names to label facets?
-m, --show-missing: Plot missing values in scatter plots as lines.
-t, --no-tril: Plot the entire square matrix, not just lower triangle.
--show-contour: Turn on contours.
--colorby: The name of a column to use as a marker variable for color.
Example:
bayeslite> .show SELECT foo, baz, quux + glorb FROM mytable
--filename myfile.png
bayeslite> .show ESTIMATE foo, baz FROM mytable_cc -g mytable_cc
"""
parser = ArgumentParser(prog='.show')
parser.add_argument('bql', type=str, nargs='+', help='BQL query')
parser.add_argument('-f', '--filename', type=str, default=None,
help='output filename')
parser.add_argument('-g', '--generator', type=str, default=None,
help='Query generator name')
parser.add_argument('-s', '--shortnames', action='store_true',
help='Use column short names?')
parser.add_argument('--no-contour', action='store_true',
help='Turn off countours (KDE).'),
parser.add_argument('--show-contour', action='store_true',
help='Turn on contours (KDE).')
parser.add_argument('-m', '--show-missing', action='store_true',
help='Plot missing values in scatterplot.')
parser.add_argument('--show-full', action='store_true')
parser.add_argument('--colorby', type=str, default=None,
help='Name of column to use as a dummy variable.')
try:
args = parser.parse_args(shlex.split(argin))
except ArgparseError as e:
self.stdout.write('%s' % (e.message,))
return
bql = " ".join(args.bql)
figure = bdbcontrib.pairplot(self._bdb, bql,
generator_name=args.generator, show_contour=args.show_contour,
colorby=args.colorby, show_missing=args.show_missing,
show_full=args.show_full)
if args.filename is None:
plt.show()
else:
figure.savefig(args.filename)
plt.close('all')
def pairplot(self, cols, plotfile=None, colorby=None, **kwargs):
"""Wrap bdbcontrib.plot_utils.pairplot to show the given columns.
Specifies bdb, query with the given columns, and generator_name:
bdbcontrib_pairplot
"""
if len(cols) < 1:
raise ValueError('Pairplot at least one variable.')
qcols = cols if colorby is None else set(cols + [colorby])
query_columns = '''"%s"''' % '''", "'''.join(qcols)
with logged_query(query_string='pairplot cols=?', bindings=(query_columns,),
name=self.session_capture_name):
self.logger.plot(plotfile,
bdbcontrib.pairplot(self.bdb, '''SELECT %s FROM %s''' %
(query_columns, self.name),
generator_name=self.generator_name,
colorby=colorby,
**kwargs))
value1 = str(table_tr[j][0]),
value2 = str(table_tr[j][1]),
value3 = str(table_tr[j][2]))
header1.append(row)
footer = str(open("html/footer_select.html").read())
header1.append(footer)
final = ' '.join(header1)
return final
#pairplot
liste = test.quick_describe_columns()
gen_cols = [str(liste.iloc[0,1]),str(liste.iloc[1,1])]
PP = bdbcontrib.pairplot(bdb, '''
SELECT {string1},{string2}
FROM dfr;'''.format(string1='"' + gen_cols[0] + '"', string2='"' + gen_cols[1] + '"'));
plt.savefig("images/PP_"+basename,bbox_extra_artists=(lgd,), bbox_inches='tight')
#Missing values
n = len(matt)
def miss():
matt = np.array(bdbcontrib.describe_generator_columns(bdb, 'dfr_cc'))
array = datareduce.columns.values
header = [str(open("html/header_table.html").read())]
for (j,k) in zip(range(0, (n/2)-1),range(n/2, n-1)):
row = str(open("html/row_table.html").read()).format(colname1 = str(matt[j][1]),
count1 = datar.count()[j],
def pairplot(self, argin):
"""plot array of plots for all pairs of columns
<query> [<options>]
- continuous-continuous as scatter (optional KDE contour)
- continuous-categorical as violinplot
- categorical-categorical as heatmap
Options:
-f, --filename: the output filename. If not specified, tries
to draw.
-g, --generator: the generator name. Providing the generator
name make help to plot more intelligently.
-s, --shortnames: Use column short names to label facets?
-m, --show-missing: Plot missing values in scatter plots as lines.
-t, --no-tril: Plot the entire square matrix, not just lower triangle.
--show-contour: Turn on contours.
--colorby: The name of a column to use as a marker variable for color.
Example:
bayeslite> .show SELECT foo, baz, quux + glorb FROM mytable
--filename myfile.png
bayeslite> .show ESTIMATE foo, baz FROM mytable_cc -g mytable_cc
"""
parser = ArgumentParser(prog='.show')
parser.add_argument('bql', type=str, nargs='+', help='BQL query')
parser.add_argument('-f',
'--filename',
type=str,
default=None,
help='output filename')
parser.add_argument('-g',
'--generator',
type=str,
default=None,
help='Query generator name')
parser.add_argument('-s',
'--shortnames',
action='store_true',
help='Use column short names?')
parser.add_argument('--no-contour',
action='store_true',
help='Turn off countours (KDE).'),
parser.add_argument('--show-contour',
action='store_true',
help='Turn on contours (KDE).')
parser.add_argument('-m',
'--show-missing',
action='store_true',
help='Plot missing values in scatterplot.')
parser.add_argument('--show-full', action='store_true')
parser.add_argument('--colorby',
type=str,
default=None,
help='Name of column to use as a dummy variable.')
try:
args = parser.parse_args(shlex.split(argin))
except ArgparseError as e:
self.stdout.write('%s' % (e.message, ))
return
bql = " ".join(args.bql)
figure = bdbcontrib.pairplot(self._bdb,
bql,
generator_name=args.generator,
show_contour=args.show_contour,
colorby=args.colorby,
show_missing=args.show_missing,
show_full=args.show_full)
if args.filename is None:
plt.show()
else:
figure.savefig(args.filename)
plt.close('all')
value2=str(table_tr[j][1]),
value3=str(table_tr[j][2]))
header1.append(row)
footer = str(open("html/footer_select.html").read())
header1.append(footer)
final = ' '.join(header1)
return final
#pairplot
liste = test.quick_describe_columns()
gen_cols = [str(liste.iloc[0, 1]), str(liste.iloc[1, 1])]
PP = bdbcontrib.pairplot(
bdb, '''
SELECT {string1},{string2}
FROM dfr;'''.format(string1='"' + gen_cols[0] + '"',
string2='"' + gen_cols[1] + '"'))
plt.savefig("images/PP_" + basename,
bbox_extra_artists=(lgd, ),
bbox_inches='tight')
#Missing values
n = len(matt)
def miss():
matt = np.array(bdbcontrib.describe_generator_columns(bdb, 'dfr_cc'))
array = datareduce.columns.values
header = [str(open("html/header_table.html").read())]
for (j, k) in zip(range(0, (n / 2) - 1), range(n / 2, n - 1)):