questions about matplotlib.dates.DateFormatter() and xticks()

Question

I have a csv file like this:

trade_dt    IC              pValue
2011/2/28   0.005834749     0.794163818
2011/3/31   -0.016185318    0.465113735
2011/4/29   0.0535994       0.014731813
2011/5/31   -0.023620263    0.278485983
2011/6/30   -0.000771114    0.971613898
2011/7/29   -0.063454359    0.003181087
2011/8/31   0.13892734      6.60E-11
2011/9/30   0.003624149     0.864731069
2011/10/31  0.055471084     0.008852478

first,I read this file using pd.read_csv.Second,I transform the trade_dt using pd.to_datetime(). When I plot using ax.xaxis.set_major_formatter(mdate.DateFormatter('%Y%m%d')),it shows, for example 2011-2-28 00:00

I have two questions:

1. how to remove the minute and second in the 2011-2-28 00:00
2. how to show the xticks every 6 month instead of show it every month

here is my code.Please ignore some code which is not related to this problem and the indent

def analysis_factor_ic(_df_ic):

corrData = _df_ic.copy()
corrData['monthly_IC'] = corrData.IC.rolling(window=5, center=False).mean()
corrData['xtick'] = range(len(corrData))

plt.rcParams['font.sans-serif'] = ['SimHei']
plt.rcParams['axes.unicode_minus'] = False
fig = plt.figure(figsize=(16, 6))
ax = fig.add_subplot(111)

sigPos = corrData[(corrData.IC > 0) & (corrData.pValue < 0.05)]
sigNeg = corrData[(corrData.IC < 0) & (corrData.pValue < 0.05)]
nonSig = corrData[~(corrData.pValue < 0.05)]

ax.bar(sigPos.xtick, sigPos.IC, align='center', color='r', label=u'显著为正')
ax.bar(nonSig.xtick, nonSig.IC, align='center', color='grey', label=u'不显著')
ax.bar(sigNeg.xtick, sigNeg.IC, align='center', color='g', label=u'显著为负')

plt.legend(loc='best')

ax.xaxis.set_major_formatter(mdate.DateFormatter('%Y%m%d'))
plt.xticks(corrData.xtick[1:-1:3], corrData.index[1:-1:3])
plt.xlim(-1, len(corrData))
fig.autofmt_xdate(rotation=-90, ha='center')

plt.grid(axis='y')
plt.title(u'因子Rank IC值')

plt.show()


df_ic = pd.read_csv('dfic.csv')
df_ic['trade_dt'] = pd.to_datetime(df_ic['trade_dt'], format='%Y-%m-%d')
df_ic.set_index(df_ic['trade_dt'], inplace=True)
analysis_factor_ic(df_ic)