How to get trendline equation after using polyfit?

Question

This is my data:

_time is on the x-axis and IdCount is y-axis. I converted _time first to datetime and then to float by using this code:

#to datetime
df['_time'] = pd.to_datetime(df['_time'])
#creating list of _time column
time = df['_time'].tolist()
#time to float in order to make np.polyfit work
def datetime_to_float(d):
    return d.timestamp()

time_in_float = []

for i in time:
    time_in_float.append(datetime_to_float(i))

After this I added time_in_float to my pandas dataframe:

df['time_float'] = time_in_float

and defined x- and y-variables:

x = df['time_float']
y = df['IdCount']

I used the following code to draw a trendline in my data:

plt.plot(x, y)

z = np.polyfit(x, y, 1)
p = np.poly1d(z)
plt.plot(x,p(x),"r--")

plt.show()

I now would like to get the equation of this trendline: y = ax + b

I used this code I found on StackOverflow:

print("{0}x + {1}".format(*z))

But the parameters I got do not make a lot of sense.

How should I modify my code?

Thanks!

Edit: The code here is okay, but with the solution provided it works better.

Answer 1

I suggest you to convert your dates to float (in seconds) this way :

df['_time'] = pd.to_datetime(df['_time'])

# compute timedelta from initial time
dt = df['_time'] - df['_time'][0]

# convert timedelta to seconds
# you can choose hours or days or other units here if you want
df['time_float']= dt.astype('timedelta64[s]')