def get_df(max_date=-1):
df = CsvL.get_avatar_info()
df = dh.add_avatar_data(df, cut_pce=[5, 6, 7])
df = dh.add_session_date(df)
df.drop(['AvatarId'], axis=1, inplace=True)
df = dm.filter_date(df, 'YMD', max_date)
df = Users.merge_users_clean(df)
# filtro colonne per primo test sensibilita
df = df.drop(['Age', 'YearMonth', 'ClientCode', 'FarmaId', 'Latitudine'],
axis=1)
return df
def get_df(max_date=-1):
df = CsvL.get_prod_history()
df = dh.add_avatar_data(df, cut_pce=[5, 6, 7])
__hist_hardfix(df)
df = dh.add_session_date(df)
df.drop(['AvatarId'], axis=1, inplace=True)
df = dm.filter_date(df, 'YMD', max_date)
df = Users.merge_users_clean(df)
df = df[[
'Id', 'UserId', 'SessionId', 'ActionType', 'ProductId', 'AvSessId',
'AvatarPce', 'YMD', 'NameSurname', 'Regione', 'RoleId', 'FarmaId',
'Latitudine', 'ProvId', 'Sex'
]]
df = dh.add_prod_name(df)
return df
def progresso(df, soglia, ax=None):
if ax is None:
f = plt.figure(figsize=(9, 8))
ax = f.add_subplot(111)
dsess_max = df.groupby('UserId')['SessionId'].max().reset_index()
urw = dh.rwcount_base(df, 'UserId', 'Product')
sessratio = pd.merge(dsess_max, urw, on='UserId')
utanti = sessratio[sessratio.SessionId >= soglia].UserId.values
upochi = sessratio[sessratio.SessionId < soglia].UserId.values
dtanti = df[df.UserId.isin(utanti)]
dpochi = df[df.UserId.isin(upochi)]
rwta = dh.rwcount_base(dtanti, 'SessionId', 'Product').reset_index()
rwpo = dh.rwcount_base(dpochi, 'SessionId', 'Product').reset_index()
rwtutti = dh.rwcount_base(df, 'SessionId', 'Product').reset_index()
rwtutti = rwtutti[rwtutti.SessionId < soglia]
l1 = ax.bar(rwta.SessionId, rwta.Ratio, zorder=1)
l2 = ax.scatter(rwpo.SessionId,
rwpo.Ratio,
zorder=3,
s=40,
facecolor=co.ab_colors['rosso'],
marker='v')
l3 = ax.scatter(rwtutti.SessionId,
rwtutti.Ratio,
marker='o',
zorder=2,
facecolor=co.ab_colors['giallo'],
s=40)
ax.legend([l1, l2, l3], [
'Almeno {0} sessioni'.format(soglia),
'Meno di {0} sessioni'.format(soglia), 'Tutti'
],
fontsize=14,
loc='lower right')
ax.set_title('Correttezza media al crescere dell\'impegno', size=20)
ax.set_ylabel('Correttezza', size=16)
ax.set_xlabel('Sessioni', size=16)
vals = [0, 0.25, 0.5, 0.75, 1]
ax.set_yticks(vals)
ax.yaxis.set_major_formatter(ticker.PercentFormatter())
ax.set_yticklabels(['{:,.0%}'.format(x) for x in vals])
def av_freq_hist(df, title="", ylbl="", legend=True):
if title == "":
title = 'Gli avatar più giocati hanno ricevuto prodotti corretti?'
if ylbl == "":
ylbl = 'Numero di prodotti consigliati'
f = plt.figure(figsize=(9, 8))
ax = f.add_subplot(111)
df = df.sort_values('nTot')
ax.barh(range(0, len(df)), df.Ratio * 100)
for i, (i_pce) in enumerate(df.AvatarPce):
i_hist = ax.get_children()[i]
i_hist.set_color(co.abc_l[i_pce - 1])
i_hist.set_height(0.8)
i_hist.set_edgecolor('k')
i_hist.set_linewidth(1)
if legend:
l_hand = []
for i in range(0, 4):
i_patch = mpatches.Patch(color=co.abc_l[i],
lw=1,
ec='k',
label=dh.pce_descr(i + 1))
l_hand.append(i_patch)
ax.legend(handles=l_hand, fontsize=16, edgecolor='k')
for i, (i_name, i_tot,
i_sess) in enumerate(zip(df.AvName, df.nTot, df.SessionId)):
i_lbl = "{0} - {1} - {2}".format(int(i_tot), i_name, i_sess)
ax.text(1, i - 0.1, i_lbl, color="k", va="center", size=14)
ax.xaxis.set_major_formatter(ticker.PercentFormatter())
__barh_ax_set(ax, title)
ax.set_xticks([0, 25, 50, 75, 100])
ax.set_yticks([])
xgrid = ax.xaxis.get_gridlines()
xgrid[1].set_color('k')
xgrid[1].set_ls('--')
xgrid[1].set_lw(1)
xgrid[2].set_color('k')
xgrid[2].set_ls('--')
xgrid[2].set_lw(1)
xgrid[3].set_color('k')
xgrid[3].set_ls('--')
xgrid[3].set_lw(1)
ax.set_xlim([0, 100])
ax.set_ylim([-0.75, len(df) - 0.25])
ax.set_xlabel(r'Correttezza', size=18)
ax.set_ylabel(ylbl, size=18)
f.tight_layout()
def plot_uhist(uid, type_suffix):
uhist = Info.get_user_history(uid)
uhistrw = rwcount(uhist, 'SessionId', type_suffix).fillna(0)
arr_start = dh.sess_start(uhist)
gm.plot_uhist(uid, uhistrw, arr_start)
def rwcount(df, group, type_suffix='Info', count_col='Id'):
return dh.rwcount_base(df, group, type_suffix, count_col)
sns.set()
sns.set_style('ticks')
sns.set_palette(Constants.abc_l)
def rwcount(df, group, type_suffix='Info', count_col='Id'):
return dh.rwcount_base(df, group, type_suffix, count_col)
df = Info.get_df() # dataframe base
# {{{ Grafico prodotti piu frequenti
if 0:
rw_info = rwcount(df, 'ProductId', "Info")
rw_info_freq = dh.add_prod_name(rw_info.nlargest(20, 'nTot'))
gm.freq_hist(rw_info_freq, "Le informazioni sui prodotti più frequenti "
"sono corrette?", "Numero di informazioni date")
if 0:
rw_info = rwcount(df, 'ProductId', "Benefit")
rw_info_freq = dh.add_prod_name(rw_info.nlargest(20, 'nTot'))
gm.freq_hist(rw_info_freq, "I benefici indicati per i prodotti "
"più frequenti sono corretti?", "Numero di benefici indicati")
if 0:
rw_info = rwcount(df, 'ProductId', "Advantages")
rw_info_freq = dh.add_prod_name(rw_info.nlargest(20, 'nTot'))
gm.freq_hist(rw_info_freq, "I vantaggi indicati per i prodotti "
"più frequenti sono corretti?", "Numero di vantaggi indicati")
# }}}