CODEWITHSUNDEEP
pythonpython-3.xpandasnumpydataframe
Sirmione
Sirmione

Reputation: 311

How to print a value from a different column if certain conditions are met in an other column?

I need to print a value from a different column if certain conditions are met in an other column.

The df consist of 4 columns year, gdp, the difference and the signal

df=pd.read_excel("file.xls", names = ["year","GDP"])
gdp["diff"] = gdp["GDP"].diff(1)
signal = gdp["diff"].clip(lower = -1.0, upper=1.0)
gdp["signal"] = signal

The signal is -1 for neg values and +1 for pos values.

The condition is I have to print the year for which there are 2 consecutive negative periods.

rec_start=(gdp["signal"]==-1) & (gdp["signal"].shift(-1)==-1)
gdp["start"]=rec_start # which gives a boolean mask

rec and start are the same 

     year      GDP   diff  signal  start
0  1999q4  12323.3    NaN     NaN  False
1  2000q1  12359.1   35.8     1.0  False
2  2000q2  12592.5  233.4     1.0  False
3  2000q3  12607.7   15.2     1.0  False
4  2000q4  12679.3   71.6     1.0  False
5  2001q1  12643.3  -36.0    -1.0  False
6  2001q2  12710.3   67.0     1.0  False
7  2001q3  12670.1  -40.2    -1.0  False
8  2001q4  12705.3   35.2     1.0  False
9  2002q1  12822.3  117.0     1.0  False

Now i just have to figure out the right sintax to print the year row for 2 consecutive Trues.

trying with

foo=gdp.ix[(gdp["signal"]==-1) & (gdp["signal"].shift(-1)==-1)]["year"].iloc[0] print(foo)

Did the trick.

Any help is much appreciated!

Upvotes: 1

Views: 125

Answers (1)

shivsn
shivsn

Reputation: 7838

IIUC this should work:

df.year[np.where((df.signal==-1),(df.signal==df.signal.shift‌​()),0).astype('bool'‌​)]

Upvotes: 1

Related Questions