def drop_hidden_codes(view):
'''
'''
#drop hidden codes
if 'x_hidden_codes' in view.meta():
vdf = helpers.deep_drop(view.dataframe,
view.meta()['x_hidden_codes'],
axes=0)
else:
vdf = view.dataframe
return vdf
def PowerPointPainter(full_template_path,
pptx_output_path,
meta,
clusters,
cut_by_these_crsbrks=['@'],
chart_type='bar',
text_key=None):
"""
Builds PPTX file from cluster, list of clusters, or dictionary of
clusters.
param: full_template_path - path to YG pptx template
param: pptx_output_path - path to where the newly generated pptx
will be stored
param: meta - metadata as dictionary used to paint
datframes
param: clusters - quantipy clusters, must be provided in a list
"""
#check cluster type, if not a list, convert to list.
if not isinstance(clusters, list):
clusters = [clusters]
#get the default text key if none provided
if text_key is None:
text_key = meta['lib']['default text']
#loop over clusters, returns pptx for each cluster.
for i, cluster in enumerate(clusters):
save_prs = False
prs = Presentation(full_template_path)
target_crossbreaks = cut_by_these_crsbrks
numofdownbreaks = len(cluster[cluster.keys()[0]].content_of_axis)
print('\nBuilding PPTX for {file_name}').format(file_name=cluster.name)
for x in range(0, numofdownbreaks):
for chk in cluster:
chain = cluster[chk]
side = chain.content_of_axis[x]
top = chain.source_name
if top in target_crossbreaks:
for idx, v in enumerate(chain.views):
view = chain[chain.data_key][chain.filter][side][top][v]
vkey = v.split('|')
weight = vkey[0]
func = vkey[1]
relation = vkey[2]
rel_to = vkey[3]
weight = vkey[4]
name = vkey[5]
#drop hidden codes
if 'x_hidden_codes' in view.meta:
vdf = helpers.deep_drop(
view.dataframe,
view.meta['x_hidden_codes'],
axes=0
)
else:
vdf = view.dataframe
#add question and value labels to df
if 'x_new_order' in view.meta:
df = helpers.paint_dataframe(
df=vdf.copy(),
meta=meta,
ridx=view.meta['x_new_order'],
text_key=text_key
)
else:
df = helpers.paint_dataframe(
df=vdf.copy(),
meta=meta,
text_key=text_key
)
#get question label
question_label = df.index[0][0]
#remove nesting
df.columns = df.columns.droplevel(0)
df.index = df.index.droplevel(0)
#rename @1 as Total
df.rename(columns = {'@1': 'Total'}, inplace=True)
#base df
if v.startswith('x|freq|x:y||') and weight!='':
base_description = view.meta['agg']['text'].strip()
base_text = get_base(
df,
base_description,
grid_summary=False
)
#percentage df
if v.startswith('x|freq||y|') and weight!='':
''' clean and round cells '''
df_table = df
df_table.columns = pd.Series(df_table.columns).str.replace('&', 'and')
df_table.index = pd.Series(df_table.index).str.replace('&', 'and')
'----SPLIT DFS & LOOP OVER THEM--------------------------------------------------'
collection_of_dfs = df_splitter(
df_table,
min_rows=5,
max_rows=15
)
for i, df_table_slice in enumerate(collection_of_dfs):
numofcols = len(df_table_slice.columns)
numofrows = len(df_table_slice.index)
df_table_slice = df_table_slice/100
'----ADDPEND SLIDE TO PRES----------------------------------------------------'
slide_layout = prs.slide_masters[0].slide_layouts[1]
slide = prs.slides.add_slide(slide_layout)
'----ADD SHAPES TO SLIDE------------------------------------------------------'
''' title shape '''
slide_title_text = "Click to add slide title"
if i > 0:
slide_title_text_cont = (
'%s (continued %s)' % (
slide_title_text,
i+1
)
)
else:
slide_title_text_cont = slide_title_text
slide_title = add_textbox(
slide,
text=slide_title_text_cont,
font_color=(0,0,0),
font_size=36,
font_bold=False,
vertical_alignment='middle',
left=284400,
top=309600,
width=8582400,
height=691200
)
''' sub title shape '''
sub_title_shp = add_textbox(
slide,
text=question_label,
font_size=12,
font_italic=True,
left=284400,
top=1007999,
width=8582400,
height=468000
)
''' chart shape '''
#single series table with less than 3 categories = pie
if numofcols == 1 and numofrows <= 3:
chart = chart_selector(
slide,
df_table_slice,
'pie',
has_legend=True
)
#handle incorrect chart type requests - pie chart cannot handle more than 1 column
elif chart_type == 'pie' and numofcols > 1:
chart = chart_selector(
slide,
df_table_slice,
chart_type,
has_legend=True
)
#chart_type='bar'
#single series table with more than, equal to 4 categories and is not a
#pie chart = chart type selected dynamically chart type with no legend
elif numofcols == 1 and chart_type != 'pie':
chart = chart_selector(
slide,
df_table_slice,
chart_type,
has_legend=False
)
else:
#multi series tables = dynamic chart type with legend
chart = chart_selector(
slide,
df_table_slice,
chart_type,
has_legend=True
)
''' footer shape '''
base_text_shape = add_textbox(
slide,
text=base_text,
font_size=8,
left=284400,
top=5652000,
width=8582400,
height=396000
)
save_prs = True
if save_prs:
prs.save('{pres_path}\\{pres_name}_({cluster_name}).pptx'.format(
pres_path=pptx_output_path,
pres_name=chain.data_key,
cluster_name=cluster.name
))
print('\nprs saved')
print('\nScript Completed')