def test_multi_index_df_to_jinja_table():
# Create multi-index table
idx = np.array([['super1', 'super1', 'super1', 'super1', 'super2', 'super2', 'super2'],
['a', 'a', 'b', 'b', 'c', 'c', 'c'],
['aa', 'aa', 'ba', 'bb', 'ca', 'cb', 'cc']])
idx_tuples = list(zip(*idx))
multi_index = pd.MultiIndex.from_tuples(idx_tuples, names=['super-level', 'a-level', 'aa-level'])
columns = ['This is an incredibly long column name', 'column2', 'column3', 'column4', 'column5']
data = pd.DataFrame(np.random.rand(7, 5) * 2 - 1, index=multi_index, columns=columns)
fmt_expand_multi_index = tf.FmtExpandMultiIndex(operator=tf.OP_SUM, bold=True,
hline_color=colors.DARK_BLUE)
fmt_nDecimal = tf.FmtDecimals(n=2)
fmt_align_cells = tf.FmtAlignCellContents(alignment='right')
fmt_heatmap_1 = tf.FmtHeatmap(columns=['column2', 'column3'], rows=['aa', 'ac'], threshold=0., axis=0)
fmt_heatmap_2 = tf.FmtHeatmap(columns=['column4', 'column5'], rows=['ca', 'cc'], threshold=0.3,
min_color=colors.PURPLE, max_color=colors.ORANGE)
fmt_stripes_bg = tf.FmtStripeBackground(first_color=colors.LIGHT_GREY)
fmt_rotate_header = tf.FmtHeader(fixed_width='500px', top_padding='200px')
formatters = [fmt_expand_multi_index, fmt_align_cells, fmt_stripes_bg, fmt_heatmap_1, fmt_heatmap_2,
fmt_rotate_header, fmt_nDecimal]
# Create table
table = Block(data, formatters=formatters)
filename = 'Multi_index_table.pdf'
table.save(filename)
# And clean up file afterwards
os.remove(filename)
def test_formatters_with_operator_on_df_with_nans_replaces():
df = pd.DataFrame(np.arange(8, dtype=float).reshape(2, 4), index=['a', 'b'], columns=['aa', 'bb', 'cc', 'dd'])
df.iloc[0, 0] = np.nan
fmt_nan_replace = tf.FmtReplaceNaN()
fmt_sum_rows = tf.FmtAppendTotalsRow()
fmt_sum_columns = tf.FmtAppendTotalsColumn()
b_no_replace = Block(df, formatters=[fmt_sum_rows, fmt_sum_columns], use_default_formatters=False)
b_replace = Block(df, formatters=[fmt_nan_replace, fmt_sum_rows, fmt_sum_columns],
use_default_formatters=False)
assert_series_equal(b_replace.df.iloc[:, -1], b_no_replace.df.iloc[:, -1], check_dtype=False)
assert_series_equal(b_replace.df.iloc[-1, :], b_no_replace.df.iloc[-1, :], check_dtype=False)
def test_column_multi_index_df_to_jinja_table():
# Create column multi-index table
df = pd.DataFrame(np.arange(12, dtype=float).reshape(3, 4), index=['a', 'b', 'c'], columns=['aa', 'bb', 'cc', 'aa'])
df['grouping'] = 'g'
df = df.reset_index()
df = df.groupby(['grouping', 'index']).sum().unstack()
# Create formatters
row_css = ['text-align:center', 'text-align:right']
index_col_css = ['background-color: #00FFFF'] * 2
fmt_mi_header = tf.FmtColumnMultiIndexRows(row_css=row_css, index_col_css=index_col_css)
# Create table
table = Block(df, formatters=[fmt_mi_header])
filename = 'Column_multi_index_table.pdf'
table.save(filename)
# And clean up file afterwards
os.remove(filename)
def test_df_to_jinja_table_add_extra_formatters():
fmt_heatmap = blformat.FmtHeatmap(threshold=0., axis=0)
fmt_add_totals_mean = blformat.FmtAppendTotalsRow(row_name='Sum', operator=blformat.OP_SUM, bold=True)
fmt_fontsize_20 = blformat.FmtFontsize(fontsize=20)
formatters = [fmt_fontsize_20, fmt_heatmap, fmt_add_totals_mean]
return Block(FLAT_DATA, formatters=formatters, use_default_formatters=False)
def get_weights_html(self):
fmt_pct = tf.FmtPercent(1, apply_to_header_and_index=False)
fmt_align = tf.FmtAlignTable("left")
fmt_background = tf.FmtStripeBackground(
first_color=tf.colors.LIGHT_GREY,
second_color=tf.colors.WHITE,
header_color=tf.colors.BLACK)
targetweights = Block(self.InputList[3],
formatters=[fmt_pct, fmt_align, fmt_background],
use_default_formatters=False)._repr_html_()
targetweights = {'targetweights_table': targetweights}
effectiveweights = Block(
self.InputList[4],
formatters=[fmt_pct, fmt_align, fmt_background],
use_default_formatters=False)._repr_html_()
effectiveweights = {'effectiveweights_table': effectiveweights}
return targetweights, effectiveweights
def test_multi_index_df_to_jinja_table():
# Create multi-index table
idx = np.array([['super1', 'super1', 'super1', 'super1', 'super2', 'super2', 'super2'],
['a', 'a', 'b', 'b', 'c', 'c', 'c'],
['aa', 'ab', 'ba', 'bb', 'ca', 'cb', 'cc']])
idx_tuples = list(zip(*idx))
multi_index = pd.MultiIndex.from_tuples(idx_tuples, names=['super-level', 'a-level', 'aa-level'])
columns = ['This is an incredibly long column name', 'column2', 'column3', 'column4', 'column5']
data = pd.DataFrame(TABLE_DATA, index=multi_index, columns=columns)
fmt_expand_multi_index = blformat.FmtExpandMultiIndex(operator=blformat.OP_SUM, bold=True,
hline_color=colors.DARK_BLUE)
fmt_ndecimal = blformat.FmtDecimals(n=2)
fmt_align_cells = blformat.FmtAlignCellContents(alignment='right')
fmt_heatmap_1 = blformat.FmtHeatmap(columns=['column2', 'column3'], rows=['aa', 'ab', 'ac'], threshold=0., axis=0)
fmt_heatmap_2 = blformat.FmtHeatmap(columns=['column4', 'column5'], rows=['ca', 'cb', 'cc'], threshold=0.3,
min_color=colors.PURPLE, max_color=colors.ORANGE)
fmt_rotate_header = blformat.FmtHeader(fixed_width='500px', top_padding='200px')
formatters = [fmt_expand_multi_index, fmt_align_cells, fmt_heatmap_1, fmt_heatmap_2,
fmt_rotate_header, fmt_ndecimal]
return Block(data, formatters=formatters, use_default_formatters=False)
def get_performance_stats_html(self):
pct_rows = [('P&L', 'Total return'), ('P&L', 'Annual return'),
('P&L', 'Annual return (asset mode)'),
('Risk-adjusted return based on Drawdown',
'Max percentage drawdown')]
dec_rows = [
('P&L', 'Starting cash'),
('P&L', 'End value'),
('Risk-adjusted return based on Drawdown', 'Max money drawdown'),
# Distribution
('Distribution moments', 'Returns volatility'),
('Distribution moments', 'Returns skewness'),
('Distribution moments', 'Returns kurtosis'),
# Risk-adjusted return based on Volatility
('Risk-adjusted return based on Volatility', 'Treynor ratio'),
('Risk-adjusted return based on Volatility', 'Sharpe ratio'),
('Risk-adjusted return based on Volatility', 'Information ratio'),
# Risk-adjusted return based on Value at Risk
('Risk-adjusted return based on Value at Risk', 'VaR'),
('Risk-adjusted return based on Value at Risk',
'Expected Shortfall'),
('Risk-adjusted return based on Value at Risk', 'Excess var'),
('Risk-adjusted return based on Value at Risk',
'Conditional sharpe ratio'),
# Risk-adjusted return based on Lower Partial Moments
('Risk-adjusted return based on Lower Partial Moments',
'Omega ratio'),
('Risk-adjusted return based on Lower Partial Moments',
'Sortino ratio'),
('Risk-adjusted return based on Lower Partial Moments',
'Kappa three ratio'),
('Risk-adjusted return based on Lower Partial Moments',
'Gain loss ratio'),
('Risk-adjusted return based on Lower Partial Moments',
'Upside potential ratio'),
# Risk-adjusted return based on Drawdown
('Risk-adjusted return based on Drawdown', 'Calmar ratio')
]
fmt_pct = tf.FmtPercent(1,
rows=pct_rows,
apply_to_header_and_index=False)
fmt_dec = tf.FmtDecimals(2,
rows=dec_rows,
apply_to_header_and_index=False)
fmt_align = tf.FmtAlignTable("left")
fmt_background = tf.FmtStripeBackground(
first_color=tf.colors.LIGHT_GREY,
second_color=tf.colors.WHITE,
header_color=tf.colors.BLACK)
fmt_multiindex = tf.FmtExpandMultiIndex(operator=tf.OP_NONE)
perf_data = Block(self.InputList[2],
formatters=[
fmt_multiindex, fmt_pct, fmt_dec, fmt_align,
fmt_background
],
use_default_formatters=False)._repr_html_()
perf_data = {'performance_table': perf_data}
return perf_data
def test_df_to_jinja_table_extra_formatters_only():
return Block(FLAT_DATA,
formatters=[blformat.FmtHeatmap(threshold=0., axis=0)],
use_default_formatters=False)
def test_df_to_jinja_table_default():
return Block(FLAT_DATA, use_default_formatters=False)