python - create a plot with all dates in pandas -

i need create plot of dataframe several columns. have dates on x-axis. how possible make on plot there dates? data shown period of once every 5 months. in columns, data small, important me visible on plot. dataframe looks this.

    date      col1        col2      col3     col4        col5        col6 20.05.2016  1091,06     12932,31       0    9343,334    23913,74      0 27.05.2016  1086,66     11845,64       0    9786,654    23913,74      0 03.06.2016  1083,04     10762,59       0    9786,654    23913,74      0 10.06.2016  1083,96     9678,630    4000    9786,654    23913,74      0 17.06.2016  1087,31     22718,40       0    9786,654    23913,74   1412 24.06.2016  1089,78     21628,62       0    9786,654    23828,96      0 01.07.2016  1083,70     20544,92       0    9749,567    23828,96      0 08.07.2016  1081,92     19463          0    9749,567    23828,96      0 ... 

my code looks this:

df.plot(figsize=(20,10), x='date', y=['col1', 'col2', 'col3', 'col4', 'col5', 'col6'])  plt.show() 

i grateful advice.

first use set_index , if need subset [] if need filter columns names:

cols = ['col1', 'col2', 'col3', 'col4', 'col5', 'col6'] df.set_index('date')[cols].plot(figsize=(20,10)) 

and columns of df omit it:

df.set_index('date').plot(figsize=(20,10)) 

but if need columns without 0 use boolean indexing loc , filter columns ne (!=) , all trues per columns:

#replace decimals , . , floats, check notice solution   df['date'] = pd.to_datetime(df['date']) df = df.set_index('date').replace(',', '.', regex=true).astype(float)  print (df.ne(0))             col1  col2   col3  col4  col5   col6 date                                             2016-05-20  true  true  false  true  true  false 2016-05-27  true  true  false  true  true  false 2016-03-06  true  true  false  true  true  false 2016-10-06  true  true   true  true  true  false 2016-06-17  true  true  false  true  true   true 2016-06-24  true  true  false  true  true  false 2016-01-07  true  true  false  true  true  false 2016-08-07  true  true  false  true  true  false  print (df.ne(0).all()) col1     true col2     true col3    false col4     true col5     true col6    false dtype: bool 

---

df = df.loc[:, df.ne(0).all()] print (df)                col1      col2      col4      col5 date                                              2016-05-20  1091.06  12932.31  9343.334  23913.74 2016-05-27  1086.66  11845.64  9786.654  23913.74 2016-03-06  1083.04  10762.59  9786.654  23913.74 2016-10-06  1083.96   9678.63  9786.654  23913.74 2016-06-17  1087.31  22718.40  9786.654  23913.74 2016-06-24  1089.78  21628.62  9786.654  23828.96 2016-01-07  1083.70  20544.92  9749.567  23828.96 2016-08-07  1081.92  19463.00  9749.567  23828.96   df.plot(figsize=(20,10)) 

notice:

there problem decimals, need parameter decimal in read_csv or replace astype used in solution above:

df = pd.read_csv('filename', index_col=['date'], decimal=',', parse_dates=['date'])  print (df)                col1      col2  col3      col4      col5  col6 date                                                          2016-05-20  1091.06  12932.31     0  9343.334  23913.74     0 2016-05-27  1086.66  11845.64     0  9786.654  23913.74     0 2016-03-06  1083.04  10762.59     0  9786.654  23913.74     0 2016-10-06  1083.96   9678.63  4000  9786.654  23913.74     0 2016-06-17  1087.31  22718.40     0  9786.654  23913.74  1412 2016-06-24  1089.78  21628.62     0  9786.654  23828.96     0 2016-01-07  1083.70  20544.92     0  9749.567  23828.96     0 2016-08-07  1081.92  19463.00     0  9749.567  23828.96     0