def get_report_structure(date_start: datetime, date_end: datetime,
sample: dict, summary: dict) -> Renderable:
"""Generate a HTML report from summary statistics and a given sample.
Args:
sample: A dict containing the samples to print.
summary: Statistics to use for the overview, variables, correlations and missing values.
Returns:
The profile report in HTML format
"""
sections = Sequence(
[
Dataset(
package=summary["package"],
date_start=date_start,
date_end=date_end,
values=summary["table"],
messages=summary["messages"],
variables=summary["variables"],
name="Overview",
anchor_id="overview",
),
Sequence(
render_variables_section(summary),
sequence_type="accordion",
name="Variables",
anchor_id="variables",
),
Sequence(
get_correlation_items(summary),
sequence_type="tabs",
name="Correlations",
anchor_id="correlations",
),
Sequence(
get_missing_items(summary),
sequence_type="tabs",
name="Missing values",
anchor_id="missing",
),
Sequence(
get_sample_items(sample),
sequence_type="list",
name="Sample",
anchor_id="sample",
),
],
name="Report",
sequence_type="sections",
)
return sections
def get_scatter_matrix(scatter_matrix):
image_format = config["plot"]["image_format"].get(str)
titems = []
for x_col, y_cols in scatter_matrix.items():
items = []
for y_col, splot in y_cols.items():
items.append(
Image(
splot,
image_format=image_format,
alt=f"{x_col} x {y_col}",
anchor_id=f"interactions_{x_col}_{y_col}",
name=y_col,
)
)
titems.append(
Sequence(
items,
sequence_type="tabs",
name=x_col,
anchor_id=f"interactions_{x_col}",
)
)
return titems
def render_generic(summary):
template_variables = {} # render_common(summary)
info = Overview(
anchor_id=summary["varid"],
warnings=summary["warnings"],
var_type="Unsupported",
var_name=summary["varname"],
)
table = Table([
{
"name": "Missing",
"value": summary["n_missing"],
"fmt": "fmt",
"class": "alert" if "n_missing" in summary["warn_fields"] else "",
},
{
"name": "Missing (%)",
"value": summary["p_missing"],
"fmt": "fmt_percent",
"class": "alert" if "p_missing" in summary["warn_fields"] else "",
},
{
"name": "Memory size",
"value": summary["memory_size"],
"fmt": "fmt_bytesize",
},
])
return {
"top": Sequence([info, table, HTML("")], sequence_type="grid"),
"bottom": None,
"ignore": "ignore",
}
def render_path_image(summary):
n_freq_table_max = config["n_freq_table_max"].get(int)
image_format = config["plot"]["image_format"].get(str)
template_variables = render_path(summary)
# Top
template_variables["top"].content["items"][0].content[
"var_type"] = "Image Path"
# Bottom
keys = {"Image shape": "image_shape", "Exif keys": "exif_keys"}
for title, key in keys.items():
template_variables["freqtable_{}".format(key)] = freq_table(
freqtable=summary["{}_counts".format(key)],
n=summary["n"],
max_number_to_print=n_freq_table_max,
)
# TODO: add dropdown to switch to specific values
exif_keys = FrequencyTable(
template_variables["freqtable_{}".format("exif_keys")],
name="Exif keys",
anchor_id="{varid}exif_frequency".format(varid=summary["varid"]),
)
template_variables["bottom"].content["items"].append(exif_keys)
image_shape_freq = FrequencyTable(
template_variables["freqtable_{}".format("image_shape")],
name="Frequency",
anchor_id="{varid}image_shape_frequency".format(
varid=summary["varid"]),
)
image_shape_scatter = Image(
scatter_series(summary["scatter_data"]),
image_format=image_format,
alt="Scatterplot of image sizes",
caption="Scatterplot of image sizes",
name="Scatter",
anchor_id="{varid}scatter".format(varid=summary["varid"]),
)
image_shape = Sequence(
[image_shape_freq, image_shape_scatter],
sequence_type="tabs",
name="Image shape",
anchor_id="{varid}image_shape".format(varid=summary["varid"]),
)
template_variables["bottom"].content["items"].append(image_shape)
return template_variables
def get_report_structure(date_start: datetime, date_end: datetime,
sample: dict, summary: dict) -> Renderable:
"""Generate a HTML report from summary statistics and a given sample.
Args:
sample: A dict containing the samples to print.
summary: Statistics to use for the overview, variables, correlations and missing values.
Returns:
The profile report in HTML format
"""
warnings = summary["messages"]
section_items = get_section_items()
section_items.append(
Sequence(
get_dataset_items(summary, date_start, date_end, warnings),
sequence_type="tabs",
name="Overview",
anchor_id="overview",
))
section_items.append(
Sequence(
render_variables_section(summary),
sequence_type="accordion",
name="Variables",
anchor_id="variables",
))
section_items.append(
Sequence(
get_scatter_matrix(summary["scatter"]),
sequence_type="tabs",
name="Interactions",
anchor_id="interactions",
))
corr = get_correlation_items(summary)
if corr is not None:
section_items.append(corr)
section_items.append(
Sequence(
get_missing_items(summary),
sequence_type="tabs",
name="Missing values",
anchor_id="missing",
))
section_items.append(
Sequence(
get_sample_items(sample),
sequence_type="list",
name="Sample",
anchor_id="sample",
))
sections = Sequence(section_items, name="Report", sequence_type="sections")
return sections
def get_dataset_overview(summary):
dataset_info = Table(
[
{
"name": "Total Number of Records",
"value": summary["table"]["n"],
"fmt": "fmt_numeric",
},
{
"name": "Total Number of Columns",
"value": summary["table"]["n_var"],
"fmt": "fmt_numeric",
},
{
"name": "Missing row cells",
"value": summary["table"]["n_cells_missing"],
"fmt": "fmt_numeric",
},
{
"name": "Missing row cells (%)",
"value": summary["table"]["p_cells_missing"],
"fmt": "fmt_percent",
},
{
"name": "Duplicate rows",
"value": summary["table"]["n_duplicates"],
"fmt": "fmt_numeric",
},
{
"name": "Duplicate rows (%)",
"value": summary["table"]["p_duplicates"],
"fmt": "fmt_percent",
},
],
name="Table statistics",
)
dataset_types = Table(
[{
"name": type_name,
"value": count,
"fmt": "fmt_numeric"
} for type_name, count in summary["table"]["types"].items()],
name="Variable types",
)
return Sequence(
[dataset_info, dataset_types],
anchor_id="dataset_overview",
name="Overview",
sequence_type="grid",
)
def render_url(summary):
n_freq_table_max = config["n_freq_table_max"].get(int)
n_obs_cat = config["vars"]["cat"]["n_obs"].get(int)
# TODO: merge with boolean/categorical
mini_freq_table_rows = freq_table(freqtable=summary["value_counts"],
n=summary["n"],
max_number_to_print=n_obs_cat)
template_variables = render_common(summary)
keys = ["scheme", "netloc", "path", "query", "fragment"]
for url_part in keys:
template_variables["freqtable_{}".format(url_part)] = freq_table(
freqtable=summary["{}_counts".format(url_part)],
n=summary["n"],
max_number_to_print=n_freq_table_max,
)
full_frequency_table = FrequencyTable(
template_variables["freq_table_rows"],
name="Full",
anchor_id="{varid}full_frequency".format(varid=summary["varid"]),
)
scheme_frequency_table = FrequencyTable(
template_variables["freqtable_scheme"],
name="Scheme",
anchor_id="{varid}scheme_frequency".format(varid=summary["varid"]),
)
netloc_frequency_table = FrequencyTable(
template_variables["freqtable_netloc"],
name="Netloc",
anchor_id="{varid}netloc_frequency".format(varid=summary["varid"]),
)
path_frequency_table = FrequencyTable(
template_variables["freqtable_path"],
name="Path",
anchor_id="{varid}path_frequency".format(varid=summary["varid"]),
)
query_frequency_table = FrequencyTable(
template_variables["freqtable_query"],
name="Query",
anchor_id="{varid}query_frequency".format(varid=summary["varid"]),
)
fragment_frequency_table = FrequencyTable(
template_variables["freqtable_fragment"],
name="Fragment",
anchor_id="{varid}fragment_frequency".format(varid=summary["varid"]),
)
items = [
full_frequency_table,
scheme_frequency_table,
netloc_frequency_table,
path_frequency_table,
query_frequency_table,
fragment_frequency_table,
]
template_variables["bottom"] = Sequence(items, sequence_type="tabs")
# Element composition
info = Overview(summary["varid"], summary["varname"], "URL",
summary["warnings"])
table = Table([
{
"name": "Distinct count",
"value": summary["n_unique"],
"fmt": "fmt"
},
{
"name": "Unique (%)",
"value": summary["p_unique"],
"fmt": "fmt_percent"
},
{
"name": "Missing",
"value": summary["n_missing"],
"fmt": "fmt"
},
{
"name": "Missing (%)",
"value": summary["p_missing"],
"fmt": "fmt_percent",
},
{
"name": "Memory size",
"value": summary["memory_size"],
"fmt": "fmt_bytesize",
},
])
fqm = FrequencyTableSmall(mini_freq_table_rows)
# TODO: settings 3,3,6
template_variables["top"] = Sequence([info, table, fqm],
sequence_type="grid")
return template_variables
def render_categorical(summary):
n_obs_cat = config["vars"]["cat"]["n_obs"].get(int)
template_variables = render_common(summary)
# TODO: merge with boolean
mini_freq_table_rows = freq_table(
freqtable=summary["value_counts"],
n=summary["count"],
max_number_to_print=n_obs_cat,
)
# Top
# Element composition
info = Overview(
summary["varid"], summary["varname"], "Categorical", summary["warnings"]
)
table = Table(
[
{
"name": "Distinct count",
"value": summary["n_unique"],
"fmt": "fmt",
"class": "alert" if "n_unique" in summary["warn_fields"] else "",
},
{
"name": "Unique (%)",
"value": summary["p_unique"],
"fmt": "fmt_percent",
"class": "alert" if "p_unique" in summary["warn_fields"] else "",
},
{
"name": "Missing",
"value": summary["n_missing"],
"fmt": "fmt",
"class": "alert" if "n_missing" in summary["warn_fields"] else "",
},
{
"name": "Missing (%)",
"value": summary["p_missing"],
"fmt": "fmt_percent",
"class": "alert" if "p_missing" in summary["warn_fields"] else "",
},
{
"name": "Memory size",
"value": summary["memory_size"],
"fmt": "fmt_bytesize",
},
]
)
fqm = FrequencyTableSmall(mini_freq_table_rows)
# TODO: settings 3,3,6
template_variables["top"] = Sequence([info, table, fqm], sequence_type="grid")
# Bottom
items = []
frequency_table = FrequencyTable(
# 'frequency_table',
template_variables["freq_table_rows"],
name="Common Values",
anchor_id="{varid}common_values".format(varid=summary["varid"]),
)
items.append(frequency_table)
check_compositions = config["vars"]["cat"]["check_composition"].get(bool)
if check_compositions:
composition = Table(
[
{
"name": "Contains chars",
"value": summary["composition"]["chars"],
"fmt": "fmt",
},
{
"name": "Contains digits",
"value": summary["composition"]["digits"],
"fmt": "fmt",
},
{
"name": "Contains whitespace",
"value": summary["composition"]["spaces"],
"fmt": "fmt",
},
{
"name": "Contains non-words",
"value": summary["composition"]["non-words"],
"fmt": "fmt",
},
],
name="Composition",
anchor_id="{varid}composition".format(varid=summary["varid"]),
)
length = Table(
[
{
"name": "Max length",
"value": summary["max_length"],
"fmt": "fmt_numeric",
},
{
"name": "Mean length",
"value": summary["mean_length"],
"fmt": "fmt_numeric",
},
{
"name": "Min length",
"value": summary["min_length"],
"fmt": "fmt_numeric",
},
],
name="Length",
anchor_id="{varid}lengthstats".format(varid=summary["varid"]),
)
tbl = Sequence(
[composition, length],
anchor_id="{varid}tbl".format(varid=summary["varid"]),
name="Composition",
sequence_type="grid",
)
items.append(tbl)
histogram_bins = 10
length = Image(
histogram(summary["length"], summary, histogram_bins),
alt="Scatter",
name="Length",
anchor_id="{varid}length".format(varid=summary["varid"]),
)
items.append(length)
template_variables["bottom"] = Sequence(
items,
sequence_type="tabs",
anchor_id="{varid}bottom".format(varid=summary["varid"]),
)
return template_variables
def get_correlation_items(summary) -> Optional[Renderable]:
"""Create the list of correlation items
Args:
summary: dict of correlations
Returns:
List of correlation items to show in the interface.
"""
items = get_items()
pearson_description = (
"The Pearson's correlation coefficient (<em>r</em>) is a measure of linear correlation "
"between two variables. It's value lies between -1 and +1, -1 indicating total negative "
"linear correlation, 0 indicating no linear correlation and 1 indicating total positive "
"linear correlation. Furthermore, <em>r</em> is invariant under separate changes in location "
"and scale of the two variables, implying that for a linear function the angle to the "
"x-axis does not affect <em>r</em>.<br /><br />To calculate <em>r</em> for two "
"variables <em>X</em> and <em>Y</em>, one divides the covariance of <em>X</em> and "
"<em>Y</em> by the product of their standard deviations. ")
spearman_description = """The Spearman's rank correlation coefficient (<em>ρ</em>) is a measure of monotonic
correlation between two variables, and is therefore better in catching nonlinear monotonic correlations than
Pearson's <em>r</em>. It's value lies between -1 and +1, -1 indicating total negative monotonic correlation,
0 indicating no monotonic correlation and 1 indicating total positive monotonic correlation.<br /><br />To
calculate <em>ρ</em> for two variables <em>X</em> and <em>Y</em>, one divides the covariance of the rank
variables of <em>X</em> and <em>Y</em> by the product of their standard deviations. """
kendall_description = """Similarly to Spearman's rank correlation coefficient, the Kendall rank correlation
coefficient (<em>τ</em>) measures ordinal association between two variables. It's value lies between -1 and +1,
-1 indicating total negative correlation, 0 indicating no correlation and 1 indicating total positive correlation.
<br /><br />To calculate <em>τ</em> for two variables <em>X</em> and <em>Y</em>, one determines the number of
concordant and discordant pairs of observations. <em>τ</em> is given by the number of concordant pairs minus the
discordant pairs divided by the total number of pairs."""
phi_k_description = """Phik (φk) is a new and practical correlation coefficient that works consistently between categorical, ordinal and interval variables, captures non-linear dependency and reverts to the Pearson correlation coefficient in case
of a bivariate normal input distribution. There is extensive documentation available <a href='https://phik.readthedocs.io/en/latest/index.html'>here</a>."""
cramers_description = """Cramér's V is an association measure for nominal random variables. The coefficient ranges from 0 to 1, with 0 indicating independence and 1 indicating perfect association.
The empirical estimators used for Cramér's V have been proved to be biased, even for large samples.
We use a bias-corrected measure that has been proposed by Bergsma in 2013 that can be found <a href='http://stats.lse.ac.uk/bergsma/pdf/cramerV3.pdf'>here</a>."""
key_to_data = {
"pearson": (-1, "Pearson's r", pearson_description),
"spearman": (-1, "Spearman's ρ", spearman_description),
"kendall": (-1, "Kendall's τ", kendall_description),
"phi_k": (0, "Phik (φk)", phi_k_description),
"cramers": (0, "Cramér's V (φc)", cramers_description),
}
image_format = config["plot"]["image_format"].get(str)
for key, item in summary["correlations"].items():
vmin, name, description = key_to_data[key]
diagram = Image(
plot.correlation_matrix(item, vmin=vmin),
image_format=image_format,
alt=name,
anchor_id=f"{key}_diagram",
name=name,
classes="correlation-diagram",
)
if len(description) > 0:
desc = HTML(
f'<div style="padding:20px" class="text-muted"><h3>{name}</h3>{description}</div>',
anchor_id=f"{key}_html",
classes="correlation-description",
)
tbl = Sequence([diagram, desc],
anchor_id=key,
name=name,
sequence_type="grid")
items.append(tbl)
else:
items.append(diagram)
corr = Sequence(
items,
sequence_type="tabs",
name="Correlations Tab",
anchor_id="correlations_tab",
)
if len(items) > 0:
btn = ToggleButton(
"Toggle correlation descriptions",
anchor_id="toggle-correlation-description",
name="Toggle correlation descriptions",
)
return Collapse(name="Correlations",
anchor_id="correlations",
button=btn,
item=corr)
else:
return None
def render_categorical(summary):
n_obs_cat = config["vars"]["cat"]["n_obs"].get(int)
image_format = config["plot"]["image_format"].get(str)
template_variables = render_common(summary)
# TODO: merge with boolean
mini_freq_table_rows = freq_table(
freqtable=summary["value_counts"],
n=summary["count"],
max_number_to_print=n_obs_cat,
)
# Top
# Element composition
info = Overview(summary["varid"], summary["varname"], "Categorical",
summary["warnings"])
table = Table([
{
"name": "Distinct count",
"value": summary["n_unique"],
"fmt": "fmt",
"class": "alert" if "n_unique" in summary["warn_fields"] else "",
},
{
"name": "Unique (%)",
"value": summary["p_unique"],
"fmt": "fmt_percent",
"class": "alert" if "p_unique" in summary["warn_fields"] else "",
},
{
"name": "Missing",
"value": summary["n_missing"],
"fmt": "fmt",
"class": "alert" if "n_missing" in summary["warn_fields"] else "",
},
{
"name": "Missing (%)",
"value": summary["p_missing"],
"fmt": "fmt_percent",
"class": "alert" if "p_missing" in summary["warn_fields"] else "",
},
{
"name": "Memory size",
"value": summary["memory_size"],
"fmt": "fmt_bytesize",
},
])
fqm = FrequencyTableSmall(mini_freq_table_rows)
# TODO: settings 3,3,6
template_variables["top"] = Sequence([info, table, fqm],
sequence_type="grid")
# Bottom
items = []
frequency_table = FrequencyTable(
template_variables["freq_table_rows"],
name="Common Values",
anchor_id="{varid}common_values".format(varid=summary["varid"]),
)
items.append(frequency_table)
check_compositions = config["vars"]["cat"]["check_composition"].get(bool)
if check_compositions:
length_table = Table(
[
{
"name": "Max length",
"value": summary["max_length"],
"fmt": "fmt_numeric",
},
{
"name": "Mean length",
"value": summary["mean_length"],
"fmt": "fmt_numeric",
},
{
"name": "Min length",
"value": summary["min_length"],
"fmt": "fmt_numeric",
},
],
name="Length",
anchor_id="{varid}lengthstats".format(varid=summary["varid"]),
)
histogram_bins = 10
length = Image(
histogram(summary["length"], summary, histogram_bins),
image_format=image_format,
alt="Scatter",
name="Length",
anchor_id="{varid}length".format(varid=summary["varid"]),
)
tbl = Sequence(
[length, length_table],
anchor_id="{varid}tbl".format(varid=summary["varid"]),
name="Length",
sequence_type="grid",
)
items.append(tbl)
n_freq_table_max = config["n_freq_table_max"].get(int)
citems = []
vc = pd.Series(summary["category_alias_values"]).value_counts()
citems.append(
FrequencyTable(
freq_table(freqtable=vc,
n=vc.sum(),
max_number_to_print=n_freq_table_max),
name="Categories",
anchor_id="{varid}category_long_values".format(
varid=summary["varid"]),
))
vc = pd.Series(summary["script_values"]).value_counts()
citems.append(
FrequencyTable(
freq_table(freqtable=vc,
n=vc.sum(),
max_number_to_print=n_freq_table_max),
name="Scripts",
anchor_id="{varid}script_values".format(
varid=summary["varid"]),
))
vc = pd.Series(summary["block_alias_values"]).value_counts()
citems.append(
FrequencyTable(
freq_table(freqtable=vc,
n=vc.sum(),
max_number_to_print=n_freq_table_max),
name="Blocks",
anchor_id="{varid}block_alias_values".format(
varid=summary["varid"]),
))
characters = Sequence(
citems,
name="Characters",
sequence_type="tabs",
anchor_id="{varid}characters".format(varid=summary["varid"]),
)
items.append(characters)
template_variables["bottom"] = Sequence(
items,
sequence_type="tabs",
anchor_id="{varid}bottom".format(varid=summary["varid"]),
)
return template_variables
def render_count(summary):
template_variables = render_common(summary)
image_format = config["plot"]["image_format"].get(str)
# Top
info = VariableInfo(
summary["varid"],
summary["varname"],
"Real number (ℝ / ℝ<sub>≥0</sub>)",
summary["warnings"],
)
table1 = Table([
{
"name": "Distinct count",
"value": summary["n_unique"],
"fmt": "fmt",
"alert": False,
},
{
"name": "Unique (%)",
"value": summary["p_unique"],
"fmt": "fmt_percent",
"alert": False,
},
{
"name": "Missing",
"value": summary["n_missing"],
"fmt": "fmt",
"alert": False,
},
{
"name": "Missing (%)",
"value": summary["p_missing"],
"fmt": "fmt_percent",
"alert": False,
},
])
table2 = Table([
{
"name": "Mean",
"value": summary["mean"],
"fmt": "fmt",
"alert": False
},
{
"name": "Minimum",
"value": summary["min"],
"fmt": "fmt",
"alert": False
},
{
"name": "Maximum",
"value": summary["max"],
"fmt": "fmt",
"alert": False
},
{
"name": "Zeros",
"value": summary["n_zeros"],
"fmt": "fmt",
"alert": False,
},
{
"name": "Zeros (%)",
"value": summary["p_zeros"],
"fmt": "fmt_percent",
"alert": False,
},
{
"name": "Memory size",
"value": summary["memory_size"],
"fmt": "fmt_bytesize",
"alert": False,
},
])
# TODO: replace with SmallImage...
mini_histo = Image(
mini_histogram(summary["histogram_data"], summary,
summary["histogram_bins"]),
image_format=image_format,
alt="Mini histogram",
)
template_variables["top"] = Sequence([info, table1, table2, mini_histo],
sequence_type="grid")
quantile_statistics = {
"name":
"Quantile statistics",
"items": [
{
"name": "Minimum",
"value": summary["min"],
"fmt": "fmt_numeric",
"alert": False,
},
{
"name": "5-th percentile",
"value": summary["quantile_5"],
"fmt": "fmt_numeric",
"alert": False,
},
{
"name": "Q1",
"value": summary["quantile_25"],
"fmt": "fmt_numeric",
"alert": False,
},
{
"name": "median",
"value": summary["quantile_50"],
"fmt": "fmt_numeric",
"alert": False,
},
{
"name": "Q3",
"value": summary["quantile_75"],
"fmt": "fmt_numeric",
"alert": False,
},
{
"name": "95-th percentile",
"value": summary["quantile_95"],
"fmt": "fmt_numeric",
"alert": False,
},
{
"name": "Maximum",
"value": summary["max"],
"fmt": "fmt_numeric",
"alert": False,
},
{
"name": "Range",
"value": summary["range"],
"fmt": "fmt_numeric",
"alert": False,
},
{
"name": "Interquartile range",
"value": summary["iqr"],
"fmt": "fmt_numeric",
"alert": False,
},
],
}
descriptive_statistics = {
"name":
"Descriptive statistics",
"items": [
{
"name": "Standard deviation",
"value": summary["std"],
"fmt": "fmt_numeric",
},
{
"name": "Coefficient of variation",
"value": summary["cv"],
"fmt": "fmt_numeric",
},
{
"name": "Kurtosis",
"value": summary["kurt"],
"fmt": "fmt_numeric"
},
{
"name": "Mean",
"value": summary["mean"],
"fmt": "fmt_numeric"
},
{
"name": "MAD",
"value": summary["mad"],
"fmt": "fmt_numeric"
},
{
"name": "Skewness",
"value": summary["skew"],
"fmt": "fmt_numeric"
},
{
"name": "Sum",
"value": summary["sum"],
"fmt": "fmt_numeric"
},
{
"name": "Variance",
"value": summary["var"],
"fmt": "fmt_numeric"
},
],
}
# TODO: Make sections data structure
# statistics = ItemRenderer(
# 'statistics',
# 'Statistics',
# 'table',
# [
# quantile_statistics,
# descriptive_statistics
# ]
# )
seqs = [
Image(
histogram(summary["histogram_data"], summary,
summary["histogram_bins"]),
image_format=image_format,
alt="Histogram",
caption="<strong>Histogram with fixed size bins</strong> (bins={})"
.format(summary["histogram_bins"]),
name="Histogram",
anchor_id="histogram",
)
]
fq = FrequencyTable(
template_variables["freq_table_rows"],
name="Common values",
anchor_id="common_values",
)
evs = Sequence(
[
FrequencyTable(
template_variables["firstn_expanded"],
name="Minimum 5 values",
anchor_id="firstn",
),
FrequencyTable(
template_variables["lastn_expanded"],
name="Maximum 5 values",
anchor_id="lastn",
),
],
sequence_type="tabs",
name="Extreme values",
anchor_id="extreme_values",
)
if "histogram_bins_bayesian_blocks" in summary:
histo_dyn = Image(
histogram(
summary["histogram_data"],
summary,
summary["histogram_bins_bayesian_blocks"],
),
image_format=image_format,
alt="Histogram",
caption=
'<strong>Histogram with variable size bins</strong> (bins={}, <a href="https://ui.adsabs.harvard.edu/abs/2013ApJ...764..167S/abstract" target="_blank">"bayesian blocks"</a> binning strategy used)'
.format(
fmt_array(summary["histogram_bins_bayesian_blocks"],
threshold=5)),
name="Dynamic Histogram",
anchor_id="dynamic_histogram",
)
seqs.append(histo_dyn)
template_variables["bottom"] = Sequence(
[
# statistics,
Sequence(seqs,
sequence_type="tabs",
name="Histogram(s)",
anchor_id="histograms"),
fq,
evs,
],
sequence_type="tabs",
anchor_id=summary["varid"],
)
return template_variables
def get_dataset_overview(summary):
dataset_info = Table(
[
{
"name": "Number of variables",
"value": summary["table"]["n_var"],
"fmt": "fmt_numeric",
},
{
"name": "Number of observations",
"value": summary["table"]["n"],
"fmt": "fmt_numeric",
},
{
"name": "Missing cells",
"value": summary["table"]["n_cells_missing"],
"fmt": "fmt_numeric",
},
{
"name": "Missing cells (%)",
"value": summary["table"]["p_cells_missing"],
"fmt": "fmt_percent",
},
{
"name": "Duplicate rows",
"value": summary["table"]["n_duplicates"],
"fmt": "fmt_numeric",
},
{
"name": "Duplicate rows (%)",
"value": summary["table"]["p_duplicates"],
"fmt": "fmt_percent",
},
{
"name": "Total size in memory",
"value": summary["table"]["memory_size"],
"fmt": "fmt_bytesize",
},
{
"name": "Average record size in memory",
"value": summary["table"]["record_size"],
"fmt": "fmt_bytesize",
},
],
name="Dataset statistics",
)
dataset_types = Table(
[{
"name": type_name,
"value": count,
"fmt": "fmt_numeric"
} for type_name, count in summary["table"]["types"].items()],
name="Variable types",
)
return Sequence(
[dataset_info, dataset_types],
anchor_id="dataset_overview",
name="Overview",
sequence_type="grid",
)
def get_correlation_items(summary) -> Optional[Renderable]:
"""Create the list of correlation items
Args:
summary: dict of correlations
Returns:
List of correlation items to show in the interface.
"""
items = get_items()
pearson_description = (
"The Pearson's correlation coefficient (<em>r</em>) is a measure of linear correlation "
"between two variables. It's value lies between -1 and +1, -1 indicating total negative "
"linear correlation, 0 indicating no linear correlation and 1 indicating total positive "
"linear correlation. Furthermore, <em>r</em> is invariant under separate changes in location "
"and scale of the two variables, implying that for a linear function the angle to the "
"x-axis does not affect <em>r</em>.<br /><br />To calculate <em>r</em> for two "
"variables <em>X</em> and <em>Y</em>, one divides the covariance of <em>X</em> and "
"<em>Y</em> by the product of their standard deviations. ")
spearman_description = """The Spearman's rank correlation coefficient (<em>ρ</em>) is a measure of monotonic
correlation between two variables, and is therefore better in catching nonlinear monotonic correlations than
Pearson's <em>r</em>. It's value lies between -1 and +1, -1 indicating total negative monotonic correlation,
0 indicating no monotonic correlation and 1 indicating total positive monotonic correlation.<br /><br />To
calculate <em>ρ</em> for two variables <em>X</em> and <em>Y</em>, one divides the covariance of the rank
variables of <em>X</em> and <em>Y</em> by the product of their standard deviations. """
kendall_description = """Similarly to Spearman's rank correlation coefficient, the Kendall rank correlation
coefficient (<em>τ</em>) measures ordinal association between two variables. It's value lies between -1 and +1,
-1 indicating total negative correlation, 0 indicating no correlation and 1 indicating total positive correlation.
<br /><br />To calculate <em>τ</em> for two variables <em>X</em> and <em>Y</em>, one determines the number of
concordant and discordant pairs of observations. <em>τ</em> is given by the number of concordant pairs minus the
discordant pairs divided by the total number of pairs."""
key_to_data = {
"pearson": (-1, "Pearson's r", pearson_description),
"spearman": (-1, "Spearman's ρ", spearman_description),
"kendall": (-1, "Kendall's τ", kendall_description),
"phi_k": (0, "Phik (φk)", ""),
"cramers": (0, "Cramér's V (φc)", ""),
"recoded": (0, "Recoded", ""),
}
image_format = config["plot"]["image_format"].get(str)
for key, item in summary["correlations"].items():
vmin, name, description = key_to_data[key]
diagram = Image(
plot.correlation_matrix(item, vmin=vmin),
image_format=image_format,
alt=name,
anchor_id="{key}_diagram".format(key=key),
name=name,
classes="correlation-diagram",
)
if len(description) > 0:
desc = HTML(
'<div style="padding:20px" class="text-muted"><h3>{name}</h3>{description}</div>'
.format(description=description, name=name),
anchor_id="{key}_html".format(key=key),
classes="correlation-description",
)
tbl = Sequence([diagram, desc],
anchor_id=key,
name=name,
sequence_type="grid")
items.append(tbl)
else:
items.append(diagram)
corr = Sequence(
items,
sequence_type="tabs",
name="Correlations Tab",
anchor_id="correlations_tab",
)
if len(items) > 0:
btn = ToggleButton(
"Toggle correlation descriptions",
anchor_id="toggle-correlation-description",
name="Toggle correlation descriptions",
)
return Collapse(name="Correlations",
anchor_id="correlations",
button=btn,
item=corr)
else:
return None
def render_complex(summary):
template_variables = {}
image_format = config["plot"]["image_format"].get(str)
# Top
info = Overview(
summary["varid"],
summary["varname"],
"Complex number (ℂ)",
summary["warnings"],
)
table1 = Table([
{
"name": "Distinct count",
"value": summary["n_unique"],
"fmt": "fmt"
},
{
"name": "Unique (%)",
"value": summary["p_unique"],
"fmt": "fmt_percent"
},
{
"name": "Missing",
"value": summary["n_missing"],
"fmt": "fmt"
},
{
"name": "Missing (%)",
"value": summary["p_missing"],
"fmt": "fmt_percent",
},
{
"name": "Memory size",
"value": summary["memory_size"],
"fmt": "fmt_bytesize",
},
])
table2 = Table([
{
"name": "Mean",
"value": summary["mean"],
"fmt": "fmt"
},
{
"name": "Minimum",
"value": summary["min"],
"fmt": "fmt"
},
{
"name": "Maximum",
"value": summary["max"],
"fmt": "fmt"
},
{
"name": "Zeros",
"value": summary["n_zeros"],
"fmt": "fmt"
},
{
"name": "Zeros (%)",
"value": summary["p_zeros"],
"fmt": "fmt_percent"
},
])
placeholder = HTML("")
template_variables["top"] = Sequence([info, table1, table2, placeholder],
sequence_type="grid")
# Bottom
items = [
Image(
scatter_complex(summary["scatter_data"]),
image_format=image_format,
alt="Scatterplot",
caption="Scatterplot in the complex plane",
name="Scatter",
anchor_id="{varid}scatter".format(varid=summary["varid"]),
)
]
bottom = Sequence(items, sequence_type="tabs", anchor_id=summary["varid"])
template_variables["bottom"] = bottom
return template_variables
def render_date(summary):
# TODO: render common?
template_variables = {}
# Top
info = Overview(summary["varid"], summary["varname"], "Date", [])
table1 = Table([
{
"name": "Distinct count",
"value": summary["n_unique"],
"fmt": "fmt"
},
{
"name": "Unique (%)",
"value": summary["p_unique"],
"fmt": "fmt_percent"
},
{
"name": "Missing",
"value": summary["n_missing"],
"fmt": "fmt"
},
{
"name": "Missing (%)",
"value": summary["p_missing"],
"fmt": "fmt_percent",
},
{
"name": "Memory size",
"value": summary["memory_size"],
"fmt": "fmt_bytesize",
},
])
table2 = Table([
{
"name": "Minimum",
"value": summary["min"],
"fmt": "fmt"
},
{
"name": "Maximum",
"value": summary["max"],
"fmt": "fmt"
},
# {'name': '', 'value': '', 'fmt': 'fmt'},
# {'name': '', 'value': '', 'fmt': 'fmt'},
# {'name': '', 'value': '', 'fmt': 'fmt'},
# {'name': '', 'value': '', 'fmt': 'fmt'},
])
mini_histo = Image(
mini_histogram(summary["histogram_data"], summary,
summary["histogram_bins"]),
"Mini histogram",
)
template_variables["top"] = Sequence([info, table1, table2, mini_histo],
sequence_type="grid")
# Bottom
bottom = Sequence(
[
Image(
histogram(summary["histogram_data"], summary,
summary["histogram_bins"]),
alt="Histogram",
caption="Histogram",
name="Histogram",
anchor_id="{varid}histogram".format(varid=summary["varid"]),
)
],
sequence_type="tabs",
anchor_id=summary["varid"],
)
template_variables["bottom"] = bottom
return template_variables
def render_date(summary):
varid = summary["varid"]
# TODO: render common?
template_variables = {}
image_format = config["plot"]["image_format"].get(str)
# Top
info = VariableInfo(
summary["varid"], summary["varname"], "Date", summary["warnings"]
)
table1 = Table(
[
{
"name": "Distinct count",
"value": summary["n_unique"],
"fmt": "fmt",
"alert": False,
},
{
"name": "Unique (%)",
"value": summary["p_unique"],
"fmt": "fmt_percent",
"alert": False,
},
{
"name": "Missing",
"value": summary["n_missing"],
"fmt": "fmt",
"alert": False,
},
{
"name": "Missing (%)",
"value": summary["p_missing"],
"fmt": "fmt_percent",
"alert": False,
},
{
"name": "Memory size",
"value": summary["memory_size"],
"fmt": "fmt_bytesize",
"alert": False,
},
]
)
table2 = Table(
[
{"name": "Minimum", "value": summary["min"], "fmt": "fmt", "alert": False},
{"name": "Maximum", "value": summary["max"], "fmt": "fmt", "alert": False},
]
)
mini_histo = Image(
mini_histogram(summary["histogram_data"], summary, summary["histogram_bins"]),
image_format=image_format,
alt="Mini histogram",
)
template_variables["top"] = Sequence(
[info, table1, table2, mini_histo], sequence_type="grid"
)
# Bottom
bottom = Sequence(
[
Image(
histogram(
summary["histogram_data"], summary, summary["histogram_bins"]
),
image_format=image_format,
alt="Histogram",
caption="Histogram",
name="Histogram",
anchor_id=f"{varid}histogram",
)
],
sequence_type="tabs",
anchor_id=summary["varid"],
)
template_variables["bottom"] = bottom
return template_variables
def render_real(summary):
varid = summary["varid"]
template_variables = render_common(summary)
image_format = config["plot"]["image_format"].get(str)
if summary["min"] >= 0:
name = "Real number (ℝ<sub>≥0</sub>)"
else:
name = "Real number (ℝ)"
# Top
info = VariableInfo(summary["varid"], summary["varname"], name,
summary["warnings"])
table1 = Table([
{
"name": "Distinct count",
"value": summary["n_unique"],
"fmt": "fmt",
"alert": "n_unique" in summary["warn_fields"],
},
{
"name": "Unique (%)",
"value": summary["p_unique"],
"fmt": "fmt_percent",
"alert": "p_unique" in summary["warn_fields"],
},
{
"name": "Missing",
"value": summary["n_missing"],
"fmt": "fmt",
"alert": "n_missing" in summary["warn_fields"],
},
{
"name": "Missing (%)",
"value": summary["p_missing"],
"fmt": "fmt_percent",
"alert": "p_missing" in summary["warn_fields"],
},
{
"name": "Infinite",
"value": summary["n_infinite"],
"fmt": "fmt",
"alert": "n_infinite" in summary["warn_fields"],
},
{
"name": "Infinite (%)",
"value": summary["p_infinite"],
"fmt": "fmt_percent",
"alert": "p_infinite" in summary["warn_fields"],
},
])
table2 = Table([
{
"name": "Mean",
"value": summary["mean"],
"fmt": "fmt",
"alert": False
},
{
"name": "Minimum",
"value": summary["min"],
"fmt": "fmt",
"alert": False
},
{
"name": "Maximum",
"value": summary["max"],
"fmt": "fmt",
"alert": False
},
{
"name": "Zeros",
"value": summary["n_zeros"],
"fmt": "fmt",
"alert": "n_zeros" in summary["warn_fields"],
},
{
"name": "Zeros (%)",
"value": summary["p_zeros"],
"fmt": "fmt_percent",
"alert": "p_zeros" in summary["warn_fields"],
},
{
"name": "Memory size",
"value": summary["memory_size"],
"fmt": "fmt_bytesize",
"alert": False,
},
])
histogram_bins = 10
# TODO: replace with SmallImage...
mini_histo = Image(
mini_histogram(summary["histogram_data"], summary, histogram_bins),
image_format=image_format,
alt="Mini histogram",
)
template_variables["top"] = Sequence([info, table1, table2, mini_histo],
sequence_type="grid")
quantile_statistics = Table(
[
{
"name": "Minimum",
"value": summary["min"],
"fmt": "fmt_numeric"
},
{
"name": "5-th percentile",
"value": summary["5%"],
"fmt": "fmt_numeric"
},
{
"name": "Q1",
"value": summary["25%"],
"fmt": "fmt_numeric"
},
{
"name": "median",
"value": summary["50%"],
"fmt": "fmt_numeric"
},
{
"name": "Q3",
"value": summary["75%"],
"fmt": "fmt_numeric"
},
{
"name": "95-th percentile",
"value": summary["95%"],
"fmt": "fmt_numeric"
},
{
"name": "Maximum",
"value": summary["max"],
"fmt": "fmt_numeric"
},
{
"name": "Range",
"value": summary["range"],
"fmt": "fmt_numeric"
},
{
"name": "Interquartile range (IQR)",
"value": summary["iqr"],
"fmt": "fmt_numeric",
},
],
name="Quantile statistics",
)
descriptive_statistics = Table(
[
{
"name": "Standard deviation",
"value": summary["std"],
"fmt": "fmt_numeric",
},
{
"name": "Coefficient of variation (CV)",
"value": summary["cv"],
"fmt": "fmt_numeric",
},
{
"name": "Kurtosis",
"value": summary["kurtosis"],
"fmt": "fmt_numeric"
},
{
"name": "Mean",
"value": summary["mean"],
"fmt": "fmt_numeric"
},
{
"name": "Median Absolute Deviation (MAD)",
"value": summary["mad"],
"fmt": "fmt_numeric",
},
{
"name": "Skewness",
"value": summary["skewness"],
"fmt": "fmt_numeric",
"class":
"alert" if "skewness" in summary["warn_fields"] else "",
},
{
"name": "Sum",
"value": summary["sum"],
"fmt": "fmt_numeric"
},
{
"name": "Variance",
"value": summary["variance"],
"fmt": "fmt_numeric"
},
],
name="Descriptive statistics",
)
statistics = Sequence(
[quantile_statistics, descriptive_statistics],
anchor_id=f"{varid}statistics",
name="Statistics",
sequence_type="grid",
)
seqs = [
Image(
histogram(summary["histogram_data"], summary, histogram_bins),
image_format=image_format,
alt="Histogram",
caption=
f"<strong>Histogram with fixed size bins</strong> (bins={histogram_bins})",
name="Histogram",
anchor_id=f"{varid}histogram",
)
]
fq = FrequencyTable(
template_variables["freq_table_rows"],
name="Common values",
anchor_id=f"{varid}common_values",
)
evs = Sequence(
[
FrequencyTable(
template_variables["firstn_expanded"],
name="Minimum 5 values",
anchor_id=f"{varid}firstn",
),
FrequencyTable(
template_variables["lastn_expanded"],
name="Maximum 5 values",
anchor_id=f"{varid}lastn",
),
],
sequence_type="tabs",
name="Extreme values",
anchor_id=f"{varid}extreme_values",
)
if "histogram_bins_bayesian_blocks" in summary:
histo_dyn = Image(
histogram(
summary["histogram_data"],
summary,
summary["histogram_bins_bayesian_blocks"],
),
image_format=image_format,
alt="Histogram",
caption=
'<strong>Histogram with variable size bins</strong> (bins={}, <a href="https://ui.adsabs.harvard.edu/abs/2013ApJ...764..167S/abstract" target="_blank">"bayesian blocks"</a> binning strategy used)'
.format(
fmt_array(summary["histogram_bins_bayesian_blocks"],
threshold=5)),
name="Dynamic Histogram",
anchor_id=f"{varid}dynamic_histogram",
)
seqs.append(histo_dyn)
template_variables["bottom"] = Sequence(
[
statistics,
Sequence(
seqs,
sequence_type="tabs",
name="Histogram(s)",
anchor_id=f"{varid}histograms",
),
fq,
evs,
],
sequence_type="tabs",
anchor_id=f"{varid}bottom",
)
return template_variables
def render_boolean(summary):
n_obs_bool = config["vars"]["bool"]["n_obs"].get(int)
# Prepare variables
template_variables = render_common(summary)
mini_freq_table_rows = freq_table(
freqtable=summary["value_counts"],
n=summary["n"],
max_number_to_print=n_obs_bool,
)
# Element composition
info = VariableInfo(
anchor_id=summary["varid"],
warnings=summary["warnings"],
var_type="Boolean",
var_name=summary["varname"],
)
table = Table([
{
"name": "Distinct count",
"value": summary["n_unique"],
"fmt": "fmt",
"alert": "n_unique" in summary["warn_fields"],
},
{
"name": "Unique (%)",
"value": summary["p_unique"],
"fmt": "fmt_percent",
"alert": "p_unique" in summary["warn_fields"],
},
{
"name": "Missing",
"value": summary["n_missing"],
"fmt": "fmt",
"alert": "n_missing" in summary["warn_fields"],
},
{
"name": "Missing (%)",
"value": summary["p_missing"],
"fmt": "fmt_percent",
"alert": "p_missing" in summary["warn_fields"],
},
{
"name": "Memory size",
"value": summary["memory_size"],
"fmt": "fmt_bytesize",
"alert": False,
},
])
fqm = FrequencyTableSmall(mini_freq_table_rows)
template_variables["top"] = Sequence([info, table, fqm],
sequence_type="grid")
freqtable = FrequencyTable(
template_variables["freq_table_rows"],
name="Frequency Table",
anchor_id="{varid}frequency_table".format(varid=summary["varid"]),
)
template_variables["bottom"] = Sequence(
[freqtable],
sequence_type="tabs",
anchor_id="{varid}bottom".format(varid=summary["varid"]),
)
return template_variables
