Added custom delimiter.
This commit is contained in:
Lasse Studion
parent
f9a9bec3f6
commit
d47fc3505b
214
app.py
214
app.py
@ -66,122 +66,128 @@ st.markdown(
|
|||||||
with Gephi and [Gephi Light](https://gephi.org/gephi-lite/).*"""
|
with Gephi and [Gephi Light](https://gephi.org/gephi-lite/).*"""
|
||||||
)
|
)
|
||||||
|
|
||||||
try:
|
#try:
|
||||||
# Print explainer.
|
# Print explainer.
|
||||||
expl = st.expander(label="More info")
|
expl = st.expander(label="More info")
|
||||||
with expl:
|
with expl:
|
||||||
st.write(info.explainer)
|
st.write(info.explainer)
|
||||||
|
|
||||||
|
|
||||||
|
|
||||||
# Ask for nodes file.
|
# Ask for nodes file.
|
||||||
csv_nodes = st.file_uploader(
|
csv_nodes = st.file_uploader(
|
||||||
label="Upload file with **nodes** (if you have one).", key="nodes", help=f'[Example]({info.node_example})'
|
label="Upload file with **nodes** (if you have one).", key="nodes", help=f'[Example]({info.node_example})'
|
||||||
)
|
)
|
||||||
|
|
||||||
# Ask for relations file.
|
# Ask for relations file.
|
||||||
csv_edges = st.file_uploader(label="Upload file with **relations**.", key="relations", help=f'[Example]({info.relations_example})')
|
csv_edges = st.file_uploader(label="Upload file with **relations**.", key="relations", help=f'[Example]({info.relations_example})')
|
||||||
|
|
||||||
col1, col2 = st.columns([1,2])
|
col1, col2 = st.columns([1,2])
|
||||||
|
|
||||||
# Chose separator
|
# Chose separator
|
||||||
with col1:
|
with col1:
|
||||||
# Set standard separator.
|
# Set standard separator.
|
||||||
st.session_state["sep"] = ','
|
st.session_state["sep"] = ','
|
||||||
|
|
||||||
# Ask for separator.
|
# Ask for separator.
|
||||||
sep = st.radio(
|
separators = {'comma ( , )': ',', 'semicolon ( ; )': ';', 'tab ( \u21E5 )': '\t', 'pipe (|)': '|', 'space ( )': ' ', '':''}
|
||||||
'Separator in your files:',
|
sep = st.radio(
|
||||||
options=['comma ( , )', 'semicolon ( ; )', 'tab ( \u21E5 )', 'pipe (|)', 'space ( )'],
|
'Separator in your files:',
|
||||||
help='What are the values in your files separated with?'
|
options=['comma ( , )', 'semicolon ( ; )', 'tab ( \u21E5 )', 'pipe (|)', 'space ( )', 'custom'],
|
||||||
)
|
help='What are the values in your files separated with?'
|
||||||
|
)
|
||||||
separators = {'comma ( , )': ',', 'semicolon ( ; )': ';', 'tab ( \u21E5 )': '\t', 'pipe (|)': '|', 'space ( )': ' '}
|
if sep == 'custom':
|
||||||
st.session_state["sep"] = separators[sep]
|
sep = st.text_input('Custom delimiter:')
|
||||||
|
separators[sep] = sep
|
||||||
# Preview file
|
|
||||||
with col2:
|
st.session_state["sep"] = separators[sep]
|
||||||
preview = st.button('Preview file.')
|
|
||||||
if preview:
|
|
||||||
st.dataframe(pd.read_csv(csv_edges, sep=st.session_state["sep"]))
|
|
||||||
|
|
||||||
|
|
||||||
files_uploaded = st.button('Done', 'files_uploaded')
|
|
||||||
|
|
||||||
if files_uploaded or 'files_already_uploaded' in st.session_state:
|
|
||||||
st.session_state['files_already_uploaded'] = True
|
|
||||||
|
|
||||||
if csv_edges == None:
|
|
||||||
st.markdown(':red[You need to upload a file with relations.]')
|
|
||||||
st.stop()
|
|
||||||
|
|
||||||
|
# Preview file
|
||||||
|
with col2:
|
||||||
|
preview = st.button('Preview file.')
|
||||||
|
if preview:
|
||||||
try:
|
try:
|
||||||
df = pd.read_csv(csv_edges, sep=st.session_state["sep"])
|
st.dataframe(pd.read_csv(csv_edges, sep=st.session_state["sep"]), use_container_width=True)
|
||||||
except pd.errors.EmptyDataError:
|
except pd.errors.ParserError:
|
||||||
st.markdown(':red[Have you chosen the right kind of separator?]')
|
st.markdown(':red[Have you selected a correct separator?]')
|
||||||
st.stop()
|
|
||||||
|
|
||||||
df.rename({'type': 'relation_type'}, inplace=True, axis=1) # 'type' can't be used as attribute.
|
|
||||||
df.columns = [i.lower() for i in df.columns] # Remove capital letters from column names.
|
|
||||||
|
|
||||||
with st.form('select_columns'):
|
|
||||||
|
|
||||||
# Find and store target column.
|
|
||||||
target = find_columns('target', df.columns.tolist())
|
|
||||||
|
|
||||||
# Find and store source column.
|
files_uploaded = st.button('Done', 'files_uploaded')
|
||||||
source = find_columns('source', df.columns.tolist())
|
|
||||||
|
|
||||||
# Confirm choise.
|
if files_uploaded or 'files_already_uploaded' in st.session_state:
|
||||||
columns_selected = st.form_submit_button("Done")
|
st.session_state['files_already_uploaded'] = True
|
||||||
|
|
||||||
|
if csv_edges == None:
|
||||||
|
st.markdown(':red[You need to upload a file with relations.]')
|
||||||
|
st.stop()
|
||||||
|
|
||||||
|
try:
|
||||||
|
df = pd.read_csv(csv_edges, sep=st.session_state["sep"])
|
||||||
|
except pd.errors.EmptyDataError:
|
||||||
|
st.markdown(':red[Have you chosen the right kind of separator?]')
|
||||||
|
st.stop()
|
||||||
|
|
||||||
if columns_selected:
|
df.rename({'type': 'relation_type'}, inplace=True, axis=1) # 'type' can't be used as attribute.
|
||||||
# Remove source and target columns from list of options.
|
df.columns = [i.lower() for i in df.columns] # Remove capital letters from column names.
|
||||||
columns = df.columns.tolist()
|
|
||||||
columns.remove(st.session_state["target"])
|
with st.form('select_columns'):
|
||||||
columns.remove(st.session_state["source"])
|
|
||||||
|
# Find and store target column.
|
||||||
|
target = find_columns('target', df.columns.tolist())
|
||||||
|
|
||||||
|
# Find and store source column.
|
||||||
|
source = find_columns('source', df.columns.tolist())
|
||||||
|
|
||||||
|
# Confirm choise.
|
||||||
|
columns_selected = st.form_submit_button("Done")
|
||||||
|
|
||||||
|
if columns_selected:
|
||||||
|
# Remove source and target columns from list of options.
|
||||||
|
columns = df.columns.tolist()
|
||||||
|
columns.remove(st.session_state["target"])
|
||||||
|
columns.remove(st.session_state["source"])
|
||||||
|
|
||||||
|
if all([st.session_state["source"] != "", st.session_state["target"] != ""]):
|
||||||
|
source = st.session_state["source"]
|
||||||
|
target = st.session_state["target"]
|
||||||
|
|
||||||
|
# Let the user chose what columns that should be included.
|
||||||
|
chosen_columns = st.multiselect(
|
||||||
|
label="Chose other columns to include.", options=columns, default=columns
|
||||||
|
)
|
||||||
|
|
||||||
|
if csv_nodes != None: # When a nodes file is uploaded.
|
||||||
|
df_nodes = pd.read_csv(csv_nodes, sep=st.session_state["sep"])
|
||||||
|
df_nodes.columns = [i.lower() for i in df_nodes.columns] # Remove capital letters from column names.
|
||||||
|
|
||||||
|
label_column = find_columns('label', df_nodes.columns.tolist())
|
||||||
|
df_nodes.set_index(label_column, inplace=True)
|
||||||
|
|
||||||
|
else: # If no node file provided.
|
||||||
|
nodes = list(set(df[source].tolist() + df[target].tolist()))
|
||||||
|
df_nodes = pd.DataFrame(
|
||||||
|
nodes, index=range(0, len(nodes)), columns=["labels"]
|
||||||
|
)
|
||||||
|
df_nodes.set_index("labels", inplace=True)
|
||||||
|
|
||||||
|
# Make empty graph.
|
||||||
|
G = nx.MultiDiGraph()
|
||||||
|
# Add nodes.
|
||||||
|
G = add_nodes(G, df_nodes)
|
||||||
|
# Add edges.
|
||||||
|
G = add_edges(
|
||||||
|
G, df, source=source, target=target, chosen_columns=chosen_columns
|
||||||
|
)
|
||||||
|
|
||||||
|
# Turn the graph into a string.
|
||||||
|
graph_text = "\n".join([line for line in nx.generate_gexf(G)])
|
||||||
|
|
||||||
if all([st.session_state["source"] != "", st.session_state["target"] != ""]):
|
# Download gexf-file.
|
||||||
source = st.session_state["source"]
|
gexf_file = "output.gexf"
|
||||||
target = st.session_state["target"]
|
st.download_button(
|
||||||
|
"Download gexf-file", graph_text, file_name=gexf_file
|
||||||
|
)
|
||||||
|
st.write('Import the file to Gephi/Gephi Light, or try [Gephisto](https://jacomyma.github.io/gephisto/) to get an idea of the network.')
|
||||||
|
|
||||||
# Let the user chose what columns that should be included.
|
# except:
|
||||||
chosen_columns = st.multiselect(
|
# st.markdown(':red[Something went wrong, please try again or [write to me](https://twitter.com/lasseedfast).]')
|
||||||
label="Chose other columns to include.", options=columns, default=columns
|
|
||||||
)
|
|
||||||
|
|
||||||
if csv_nodes != None: # When a nodes file is uploaded.
|
|
||||||
df_nodes = pd.read_csv(csv_nodes, sep=st.session_state["sep"])
|
|
||||||
df_nodes.columns = [i.lower() for i in df_nodes.columns] # Remove capital letters from column names.
|
|
||||||
|
|
||||||
label_column = find_columns('label', df_nodes.columns.tolist())
|
|
||||||
df_nodes.set_index(label_column, inplace=True)
|
|
||||||
|
|
||||||
else: # If no node file provided.
|
|
||||||
nodes = list(set(df[source].tolist() + df[target].tolist()))
|
|
||||||
df_nodes = pd.DataFrame(
|
|
||||||
nodes, index=range(0, len(nodes)), columns=["labels"]
|
|
||||||
)
|
|
||||||
df_nodes.set_index("labels", inplace=True)
|
|
||||||
|
|
||||||
# Make empty graph.
|
|
||||||
G = nx.MultiDiGraph()
|
|
||||||
# Add nodes.
|
|
||||||
G = add_nodes(G, df_nodes)
|
|
||||||
# Add edges.
|
|
||||||
G = add_edges(
|
|
||||||
G, df, source=source, target=target, chosen_columns=chosen_columns
|
|
||||||
)
|
|
||||||
|
|
||||||
# Turn the graph into a string.
|
|
||||||
graph_text = "\n".join([line for line in nx.generate_gexf(G)])
|
|
||||||
|
|
||||||
# Download gexf-file.
|
|
||||||
gexf_file = "output.gexf"
|
|
||||||
st.download_button(
|
|
||||||
"Download gexf-file", graph_text, file_name=gexf_file
|
|
||||||
)
|
|
||||||
st.write('Import the file to Gephi/Gephi Light, or try [Gephisto](https://jacomyma.github.io/gephisto/) to get an idea of the network.')
|
|
||||||
|
|
||||||
except:
|
|
||||||
st.markdown(':red[Something went wrong, please try again or [write to me](https://twitter.com/lasseedfast).]')
|
|
||||||
Loading…
x
Reference in New Issue
Block a user