[Carlo Lazzaro]

Joseph:
as per previos advice (that echoes FAQ), please post what you typed and what Stata gave you back via CODE delimiters (better along with an example via -dataex- command).
A graph is not enough to comment positively on the global matter,

[Joseph Welford]

* Example generated by -dataex-. To install: ssc install dataex
clear
input float Data
33
35
37
38
40
42
44
43
41
39
36
34
32
31
30
29
28
19
 9
 7
 5
 3
 1
 2
 4
 6
 8
11
14
18
20
22
25
27
26
24
23
21
17
15
13
10
12
16
end

This is the output the dataex commands gave me

[Carlo Lazzaro]

Joseph:
unfortunately I can't find at least a second variable in your -dataex-.
If your dataset is not confidential, you would be probably better off with attaching it in .dta format.

[Nick Cox]

rvplot means what? rvfplot?

So log4165, presumably the logarithm of something, is itself between about 16 and 22 million? Sounds bizarre.

When the residuals are of the order of 100s and if the response is of order of 20 million, then the model is fitting the data very well, but we don't believe a fit that good.

As others are emphasising we need to see the data used in the regression.