def bar(self, var: str, title: str = '', label: str = '') -> object:
"""
This method requires a character column (use the contents method to see column types)
and generates a bar chart.
:param var: the CHAR variable (column) you want to plot
:param title: an optional title for the chart
:param label: LegendLABEL= value for sgplot
:return: graphic plot
"""
code = "proc sgplot data=" + self.libref + '.' + self.table + self._dsopts(
)
code += ";\n\tvbar " + var
if len(label) > 0:
code += " / LegendLABEL='" + label + "'"
code += ";\n"
if len(title) > 0:
code += '\ttitle "' + title + '";\n'
code += 'run;\ntitle;'
if self.sas.nosub:
print(code)
return
ll = self._is_valid()
if not ll:
html = self.HTML
self.HTML = 1
ll = self.sas._io.submit(code)
self.HTML = html
if not self.sas.batch:
DISPLAY(HTML(ll['LST']))
else:
return ll
def columnInfo(self):
"""
display metadata about the table, size, number of rows, columns and their data type
"""
code = "proc contents data=" + self.libref + '.' + self.table + ' ' + self._dsopts(
) + ";ods select Variables;run;"
if self.sas.nosub:
print(code)
return
if self.results.upper() == 'PANDAS':
code = "proc contents data=%s.%s %s ;ods output Variables=work._variables ;run;" % (
self.libref, self.table, self._dsopts())
pd = self._returnPD(code, '_variables')
pd['Type'] = pd['Type'].str.rstrip()
return pd
else:
ll = self._is_valid()
if self.HTML:
if not ll:
ll = self.sas._io.submit(code)
if not self.sas.batch:
DISPLAY(HTML(ll['LST']))
else:
return ll
else:
if not ll:
ll = self.sas._io.submit(code, "text")
if not self.sas.batch:
print(ll['LST'])
else:
return ll
def hist(self, var: str, title: str = '', label: str = '') -> object:
"""
This method requires a numeric column (use the contents method to see column types) and generates a histogram.
:param var: the NUMERIC variable (column) you want to plot
:param title: an optional Title for the chart
:param label: LegendLABEL= value for sgplot
:return:
"""
code = "proc sgplot data=" + self.libref + '.' + self.table + self._dsopts(
)
code += ";\n\thistogram " + var + " / scale=count"
if len(label) > 0:
code += " LegendLABEL='" + label + "'"
code += ";\n"
if len(title) > 0:
code += '\ttitle "' + title + '";\n'
code += "\tdensity " + var + ';\nrun;\n' + 'title;'
if self.sas.nosub:
print(code)
return
ll = self._is_valid()
if not ll:
html = self.HTML
self.HTML = 1
ll = self.sas._io.submit(code)
self.HTML = html
if not self.sas.batch:
DISPLAY(HTML(ll['LST']))
else:
return ll
def top(self,
var: str,
n: int = 10,
order: str = 'freq',
title: str = '') -> object:
"""
Return the most commonly occuring items (levels)
:param var: the CHAR variable (column) you want to count
:param n: the top N to be displayed (defaults to 10)
:param order: default to most common use order='data' to get then in alphbetic order
:param title: an optional Title for the chart
:return: Data Table
"""
code = "proc freq data=%s.%s %s order=%s noprint;" % (
self.libref, self.table, self._dsopts(), order)
code += "\n\ttables %s / out=tmpFreqOut;" % var
code += "\nrun;"
if len(title) > 0:
code += '\ttitle "' + title + '";\n'
code += "proc print data=tmpFreqOut(obs=%s); \nrun;" % n
code += 'title;'
if self.sas.nosub:
print(code)
return
ll = self._is_valid()
if self.results.upper() == 'PANDAS':
code = "proc freq data=%s.%s%s order=%s noprint;" % (
self.libref, self.table, self._dsopts(), order)
code += "\n\ttables %s / out=tmpFreqOut;" % var
code += "\nrun;"
code += "\ndata tmpFreqOut; set tmpFreqOut(obs=%s); run;" % n
return self._returnPD(code, 'tmpFreqOut')
else:
if self.HTML:
if not ll:
ll = self.sas._io.submit(code)
if not self.sas.batch:
DISPLAY(HTML(ll['LST']))
else:
return ll
else:
if not ll:
ll = self.sas._io.submit(code, "text")
if not self.sas.batch:
print(ll['LST'])
else:
return ll
def heatmap(self,
x: str,
y: str,
options: str = '',
title: str = '',
label: str = '') -> object:
"""
Documentation link: http://support.sas.com/documentation/cdl/en/grstatproc/67909/HTML/default/viewer.htm#n0w12m4cn1j5c6n12ak64u1rys4w.htm
:param x: x variable
:param y: y variable
:param options: display options (string)
:param title: graph title
:param label:
:return:
"""
code = "proc sgplot data=%s.%s %s;" % (self.libref, self.table,
self._dsopts())
if len(options):
code += "\n\theatmap x=%s y=%s / %s;" % (x, y, options)
else:
code += "\n\theatmap x=%s y=%s;" % (x, y)
if len(label) > 0:
code += " LegendLABEL='" + label + "'"
code += ";\n"
if len(title) > 0:
code += "\ttitle '%s';\n" % title
code += "run;\ntitle;"
if self.sas.nosub:
print(code)
return
ll = self._is_valid()
if not ll:
html = self.HTML
self.HTML = 1
ll = self.sas._io.submit(code)
self.HTML = html
if not self.sas.batch:
DISPLAY(HTML(ll['LST']))
else:
return ll
def score(self,
file: str = '',
code: str = '',
out: 'SASdata' = None) -> 'SASdata':
"""
This method is meant to update a SAS Data object with a model score file.
:param file: a file reference to the SAS score code
:param code: a string of the valid SAS score code
:param out: Where to the write the file. Defaults to update in place
:return: The Scored SAS Data object.
"""
if out is not None:
outTable = out.table
outLibref = out.libref
else:
outTable = self.table
outLibref = self.libref
codestr = code
code = "data %s.%s%s;" % (outLibref, outTable, self._dsopts())
code += "set %s.%s%s;" % (self.libref, self.table, self._dsopts())
if len(file) > 0:
code += '%%include "%s";' % file
else:
code += "%s;" % codestr
code += "run;"
if self.sas.nosub:
print(code)
return None
ll = self._is_valid()
if not ll:
html = self.HTML
self.HTML = 1
ll = self.sas._io.submit(code)
self.HTML = html
if not self.sas.batch:
DISPLAY(HTML(ll['LST']))
else:
return ll
def contents(self):
"""
display metadata about the table. size, number of rows, columns and their data type ...
:return: output
"""
code = "proc contents data=" + self.libref + '.' + self.table + self._dsopts(
) + ";run;"
if self.sas.nosub:
print(code)
return
ll = self._is_valid()
if self.results.upper() == 'PANDAS':
code = "proc contents data=%s.%s %s ;" % (self.libref, self.table,
self._dsopts())
code += "ods output Attributes=work._attributes;"
code += "ods output EngineHost=work._EngineHost;"
code += "ods output Variables=work._Variables;"
code += "ods output Sortedby=work._Sortedby;"
code += "run;"
return self._returnPD(
code,
['_attributes', '_EngineHost', '_Variables', '_Sortedby'])
else:
if self.HTML:
if not ll:
ll = self.sas._io.submit(code)
if not self.sas.batch:
DISPLAY(HTML(ll['LST']))
else:
return ll
else:
if not ll:
ll = self.sas._io.submit(code, "text")
if not self.sas.batch:
print(ll['LST'])
else:
return ll
def scatter(self, x: str, y: list, title: str = '') -> object:
"""
This method plots a scatter of x,y coordinates. You can provide a list of y columns for multiple line plots.
:param x: the x axis variable; generally a time or continuous variable.
:param y: the y axis variable(s), you can specify a single column or a list of columns
:param title: an optional Title for the chart
:return: graph object
"""
code = "proc sgplot data=" + self.libref + '.' + self.table + self._dsopts(
) + ";\n"
if len(title) > 0:
code += '\ttitle "' + title + '";\n'
if isinstance(y, list):
num = len(y)
else:
num = 1
y = [y]
for i in range(num):
code += "\tscatter x=" + x + " y=" + y[i] + ";\n"
code += 'run;\n' + 'title;'
if self.sas.nosub:
print(code)
return
ll = self._is_valid()
if not ll:
html = self.HTML
self.HTML = 1
ll = self.sas._io.submit(code)
self.HTML = html
if not self.sas.batch:
DISPLAY(HTML(ll['LST']))
else:
return ll
def means(self):
"""
display descriptive statistics for the table; summary statistics. This is an alias for 'describe'
:return:
"""
dsopts = self._dsopts().partition(';\n\tformat')
code = "proc means data=" + self.libref + '.' + self.table + dsopts[
0] + " stackodsoutput n nmiss median mean std min p25 p50 p75 max;"
code += dsopts[1] + dsopts[2] + "run;"
if self.sas.nosub:
print(code)
return
ll = self._is_valid()
if self.results.upper() == 'PANDAS':
code = "proc means data=%s.%s %s stackodsoutput n nmiss median mean std min p25 p50 p75 max; %s ods output Summary=work._summary; run;" % (
self.libref, self.table, dsopts[0], dsopts[1] + dsopts[2])
return self._returnPD(code, '_summary')
else:
if self.HTML:
if not ll:
ll = self.sas._io.submit(code)
if not self.sas.batch:
DISPLAY(HTML(ll['LST']))
else:
return ll
else:
if not ll:
ll = self.sas._io.submit(code, "text")
if not self.sas.batch:
print(ll['LST'])
else:
return ll
def head(self, obs=5):
"""
display the first n rows of a table
:param obs: the number of rows of the table that you want to display. The default is 5
:return:
"""
topts = dict(self.dsopts)
topts['obs'] = obs
code = "proc print data=" + self.libref + '.' + self.table + self.sas._dsopts(
topts) + ";run;"
if self.sas.nosub:
print(code)
return
if self.results.upper() == 'PANDAS':
code = "data _head ; set %s.%s %s; run;" % (
self.libref, self.table, self.sas._dsopts(topts))
return self._returnPD(code, '_head')
else:
ll = self._is_valid()
if self.HTML:
if not ll:
ll = self.sas._io.submit(code)
if not self.sas.batch:
DISPLAY(HTML(ll['LST']))
else:
return ll
else:
if not ll:
ll = self.sas._io.submit(code, "text")
if not self.sas.batch:
print(ll['LST'])
else:
return ll
def execute_table(self, _output_type, **kwargs: dict) -> 'SASresults':
"""
executes a PROC TABULATE statement
You must specify an output type to use this method, of 'HTML', 'text', or 'Pandas'.
There are three convenience functions for generating specific output; see:
.text_table()
.table()
.to_dataframe()
:param _output_type: style of output to use
:param left: the query for the left side of the table
:param top: the query for the top of the table
:return:
"""
left = kwargs.pop('left', None)
top = kwargs.pop('top', None)
sets = dict(classes=set(), vars=set())
left._gather(sets)
if top: top._gather(sets)
table = top \
and '%s, %s' % (str(left), str(top)) \
or str(left)
proc_kwargs = dict(cls=' '.join(sets['classes']),
var=' '.join(sets['vars']),
table=table)
# permit additional valid options if passed; for now, just 'where'
proc_kwargs.update(kwargs)
# we can't easily use the SASProcCommons approach for submiting,
# since this is merely an output / display proc for us;
# but we can at least use it to check valid options in the canonical saspy way
required_options = {'cls', 'var', 'table'}
allowed_options = {'cls', 'var', 'table', 'where'}
verifiedKwargs = SASProcCommons._stmt_check(self, required_options,
allowed_options,
proc_kwargs)
if (_output_type == 'Pandas'):
# for pandas, use the out= directive
code = "proc tabulate data=%s.%s %s out=temptab;\n" % (
self.data.libref, self.data.table, self.data._dsopts())
else:
code = "proc tabulate data=%s.%s %s;\n" % (
self.data.libref, self.data.table, self.data._dsopts())
# build the code
for arg, value in verifiedKwargs.items():
code += " %s %s;\n" % (arg == 'cls' and 'class' or arg, value)
code += "run;"
# teach_me_SAS
if self.sas.nosub:
print(code)
return
# submit the code
ll = self.data._is_valid()
if _output_type == 'HTML':
if not ll:
html = self.data.HTML
self.data.HTML = 1
ll = self.sas._io.submit(code)
self.data.HTML = html
if not self.sas.batch:
DISPLAY(HTML(ll['LST']))
check, errorMsg = self.data._checkLogForError(ll['LOG'])
if not check:
raise ValueError("Internal code execution failed: " +
errorMsg)
else:
return ll
elif _output_type == 'text':
if not ll:
html = self.data.HTML
self.data.HTML = 1
ll = self.sas._io.submit(code, 'text')
self.data.HTML = html
print(ll['LST'])
return
elif _output_type == 'Pandas':
return self.to_nested_dataframe(code)
def tail(self, obs=5):
"""
display the last n rows of a table
:param obs: the number of rows of the table that you want to display. The default is 5
:return:
"""
code = "proc sql;select count(*) format best32. into :lastobs from " + self.libref + '.' + self.table + self._dsopts(
) + ";%put lastobs=&lastobs tom;quit;"
nosub = self.sas.nosub
self.sas.nosub = False
le = self._is_valid()
if not le:
ll = self.sas.submit(code, "text")
lastobs = ll['LOG'].rpartition("lastobs=")
lastobs = lastobs[2].partition(" tom")
lastobs = int(lastobs[0])
else:
lastobs = obs
firstobs = lastobs - (obs - 1)
if firstobs < 1:
firstobs = 1
topts = dict(self.dsopts)
topts['obs'] = lastobs
topts['firstobs'] = firstobs
code = "proc print data=" + self.libref + '.' + self.table + self.sas._dsopts(
topts) + ";run;"
self.sas.nosub = nosub
if self.sas.nosub:
print(code)
return
if self.results.upper() == 'PANDAS':
code = "data _tail ; set %s.%s %s; run;" % (
self.libref, self.table, self.sas._dsopts(topts))
return self._returnPD(code, '_tail')
else:
if self.HTML:
if not le:
ll = self.sas._io.submit(code)
else:
ll = le
if not self.sas.batch:
DISPLAY(HTML(ll['LST']))
else:
return ll
else:
if not le:
ll = self.sas._io.submit(code, "text")
else:
ll = le
if not self.sas.batch:
print(ll['LST'])
else:
return ll
