Skip to content

Commit

Permalink
Now filters qm vs playoffs
Browse files Browse the repository at this point in the history
  • Loading branch information
@justinb4003
justinb4003 committed Mar 23, 2024
1 parent 4773db8 commit 19d19e3
Showing 1 changed file with 124 additions and 111 deletions.
235 changes: 124 additions & 111 deletions data-streamlit/pages/match_breakdowns.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -30,9 +30,20 @@
# st.dataframe(event)

matches = load_matches_data(ek)
types = matches.comp_level.unique()
team_data = load_team_data(ek)

print(team_data.columns)
qm_filter = st.checkbox('Qualifying Matches', True)
po_filter = st.checkbox('Playoff Matches', True)

match_types = []
if qm_filter:
match_types.append('qm')
if po_filter:
for level in ['sf', 'f']:
match_types.append(level)
print(types)
print(match_types)
team_filter = st.multiselect('Team', team_data, placeholder='Select a team')

try:
Expand All @@ -41,117 +52,119 @@
except urllib.error.HTTPError:
oprdata = None
opr_totalpoints = None
matches = matches[matches['comp_level'] == 'qm']
matches = matches[matches['comp_level'].isin(match_types)]
# Order the matches dataframe by the match_number column
matches = matches.sort_values(by='match_number').reset_index(drop=True)
# st.dataframe(matches)

n1 = pd.json_normalize(matches['alliances'])
bluen2 = pd.json_normalize(n1['blue.team_keys'])
redn2 = pd.json_normalize(n1['red.team_keys'])


blueteams = pd.DataFrame(n1['blue.team_keys'].explode())
redteams = pd.DataFrame(n1['red.team_keys'].explode())
blueteams['blue.team_keys'] = blueteams['blue.team_keys'].str.replace('frc', '').astype(int)
redteams['red.team_keys'] = redteams['red.team_keys'].str.replace('frc', '').astype(int)
blueteams['robot_num'] = blueteams.groupby(blueteams.index).rank()
redteams['robot_num'] = redteams.groupby(redteams.index).rank()
pblue = blueteams.pivot(columns='robot_num', values='blue.team_keys')
pred = redteams.pivot(columns='robot_num', values='red.team_keys')
pblue.columns = ['blue1', 'blue2', 'blue3']
pred.columns = ['red1', 'red2', 'red3']
matches = matches.join(pblue).join(pred)

if team_filter:
print(team_filter)
matches = matches[(matches['red1'].isin(team_filter)) |
(matches['red2'].isin(team_filter)) |
(matches['red3'].isin(team_filter)) |
(matches['blue1'].isin(team_filter)) |
(matches['blue2'].isin(team_filter)) |
(matches['blue3'].isin(team_filter))]

if opr_totalpoints is not None:
for color in ['blue', 'red']:
for index in [1, 2, 3]:
matches = matches.join(
opr_totalpoints.set_index('teamNumber'),
on=f'{color}{index}'
)
matches.rename(
columns={'totalPoints': f'{color}{index}_totalPoints'},
inplace=True
)
# st.dataframe(matches)

statbotics = load_statbot_matches_data(get_event_key())
preds = statbotics[['comp_level', 'match_number', 'pred']]
preds = preds[preds['comp_level'] == 'qm']
pred_detail = preds.join(pd.json_normalize(preds['pred']))

actual = statbotics[['comp_level', 'match_number', 'result']]
actual = actual[actual['comp_level'] == 'qm']
actual_detail = actual.join(pd.json_normalize(actual['result']))
# st.dataframe(pred_detail)

matches = matches.join(pred_detail.set_index('match_number')
.add_suffix('_pred'),
on='match_number')

matches = matches.join(actual_detail.set_index('match_number')
.add_suffix('_actual'),
on='match_number')


for panda_idx, match in matches.iterrows():
with st.container(border=8):
# Display the match number
st.subheader(f"Match {match['match_number']}")
red, blue = st.columns(2)
stats = {}
for container, color in zip([red, blue], ['red', 'blue']):
with container:
with st.container(border=True):
alliance_opr_total = 0
team_nums = []
total_point_oprs = []
for bot in [1, 2, 3]:
team_num = match[f'{color}{bot}']
team_nums.append(team_num)
if opr_totalpoints is not None:
opr_total = match[f'{color}{bot}_totalPoints']
else:
opr_total = 0
total_point_oprs.append(opr_total)
alliance_opr_total += opr_total
pred_winner = match[f'winner_pred'] == color
actual_winner = match[f'winner_actual'] == color
redpct = match[f'red_win_prob_pred'] * 100
pct = redpct if color == 'red' else 100 - redpct
show_star = '*' if actual_winner else ''
st.markdown(f"""
### {show_star}{color.capitalize()} ({pct:.1f}%)
""")
team_breakdown = pd.DataFrame({
'Team': team_nums,
'TP OPR': total_point_oprs
}).set_index('Team')
st.table(team_breakdown.style.format(precision=1))
score_pred = match[f'{color}_score_pred']
score_actual = match[f'{color}_score_actual']
stats[color] = [
alliance_opr_total,
score_pred,
score_actual
]
# JJB: I'm not thrilled with my indent level above
misc_breakdown = pd.DataFrame({
'Label': ['Total TP OPR',
'Score Predicted',
'Score Actual'],
'Red': stats['red'],
'Blue': stats['blue'],
}).set_index('Label')
st.table(misc_breakdown.style.format(precision=1)
.highlight_max(axis='columns'))
if len(matches.index) == 0:
st.subheader('No match data available yet.')
else:
n1 = pd.json_normalize(matches['alliances'])
bluen2 = pd.json_normalize(n1['blue.team_keys'])
redn2 = pd.json_normalize(n1['red.team_keys'])

blueteams = pd.DataFrame(n1['blue.team_keys'].explode())
redteams = pd.DataFrame(n1['red.team_keys'].explode())
blueteams['blue.team_keys'] = blueteams['blue.team_keys'].str.replace('frc', '').astype(int)
redteams['red.team_keys'] = redteams['red.team_keys'].str.replace('frc', '').astype(int)
blueteams['robot_num'] = blueteams.groupby(blueteams.index).rank()
redteams['robot_num'] = redteams.groupby(redteams.index).rank()
pblue = blueteams.pivot(columns='robot_num', values='blue.team_keys')
pred = redteams.pivot(columns='robot_num', values='red.team_keys')
pblue.columns = ['blue1', 'blue2', 'blue3']
pred.columns = ['red1', 'red2', 'red3']
matches = matches.join(pblue).join(pred)

if team_filter:
print(team_filter)
matches = matches[(matches['red1'].isin(team_filter)) |
(matches['red2'].isin(team_filter)) |
(matches['red3'].isin(team_filter)) |
(matches['blue1'].isin(team_filter)) |
(matches['blue2'].isin(team_filter)) |
(matches['blue3'].isin(team_filter))]

if opr_totalpoints is not None:
for color in ['blue', 'red']:
for index in [1, 2, 3]:
matches = matches.join(
opr_totalpoints.set_index('teamNumber'),
on=f'{color}{index}'
)
matches.rename(
columns={'totalPoints': f'{color}{index}_totalPoints'},
inplace=True
)
# st.dataframe(matches)

statbotics = load_statbot_matches_data(get_event_key())
preds = statbotics[['comp_level', 'match_number', 'pred']]
preds = preds[preds['comp_level'] == 'qm']
pred_detail = preds.join(pd.json_normalize(preds['pred']))

actual = statbotics[['comp_level', 'match_number', 'result']]
actual = actual[actual['comp_level'] == 'qm']
actual_detail = actual.join(pd.json_normalize(actual['result']))
# st.dataframe(pred_detail)

matches = matches.join(pred_detail.set_index('match_number')
.add_suffix('_pred'),
on='match_number')

matches = matches.join(actual_detail.set_index('match_number')
.add_suffix('_actual'),
on='match_number')


for panda_idx, match in matches.iterrows():
with st.container(border=8):
# Display the match number
st.subheader(f"Match {match['match_number']}")
red, blue = st.columns(2)
stats = {}
for container, color in zip([red, blue], ['red', 'blue']):
with container:
with st.container(border=True):
alliance_opr_total = 0
team_nums = []
total_point_oprs = []
for bot in [1, 2, 3]:
team_num = match[f'{color}{bot}']
team_nums.append(team_num)
if opr_totalpoints is not None:
opr_total = match[f'{color}{bot}_totalPoints']
else:
opr_total = 0
total_point_oprs.append(opr_total)
alliance_opr_total += opr_total
pred_winner = match[f'winner_pred'] == color
actual_winner = match[f'winner_actual'] == color
redpct = match[f'red_win_prob_pred'] * 100
pct = redpct if color == 'red' else 100 - redpct
show_star = '*' if actual_winner else ''
st.markdown(f"""
### {show_star}{color.capitalize()} ({pct:.1f}%)
""")
team_breakdown = pd.DataFrame({
'Team': team_nums,
'TP OPR': total_point_oprs
}).set_index('Team')
st.table(team_breakdown.style.format(precision=1))
score_pred = match[f'{color}_score_pred']
score_actual = match[f'{color}_score_actual']
stats[color] = [
alliance_opr_total,
score_pred,
score_actual
]
# JJB: I'm not thrilled with my indent level above
misc_breakdown = pd.DataFrame({
'Label': ['Total TP OPR',
'Score Predicted',
'Score Actual'],
'Red': stats['red'],
'Blue': stats['blue'],
}).set_index('Label')
st.table(misc_breakdown.style.format(precision=1)
.highlight_max(axis='columns'))

0 comments on commit 19d19e3

Please sign in to comment.